PH热榜 | 2026-02-03

@alexd Thanks so much Alex for the upvote :)

@baptistejamin  @alexd Hey congrats! Curious what the biggest architectural shift was to make Hugo “agent-native” rather than just another chatbot, especially around autonomy, guardrails, and knowing when to hand off to a human without hurting CX? Rebuilding from scratch 10 years after Crisp is a bold move :)

@baptistejamin  @alexd This is solid! Will check it out as we're ramping up our support for @Pretty Prompt (25k+ users). Be great to chat with any of you guys!

@baptistejamin  @valeriansaliou  @eliottvincent such an amazing work done by the entire team!

@clementchampau Thank you man 🫶🏼

@clementchampau thanks!!!

@crgturo That's amazing! Thank you for sharing feedback and powering up your support with Hugo!

@pierre_gd Congrats for the amazing work you've done Pierre!

@thomascochet Thank you very much for supporting us for such a long time!

@osama_jaber1 Thank you!

@corey_haines Thanks, if you know you know 👀

@bahaeddin_recovery Thanks! We have a pay-as-you-go credits system that lets you set whatever monthly limit you want. AI will stop in case the limit is reached. No bad surprises, you're in control :)

@mrcalexandre Thank you very much Alex 🫶 also for the video, it's Leo's work, you know how good he is with video! Master at work

@mrcalexandre Thank you Alexandre 🙏

@zeiki_yu Yeah; we strive at making customer support experiences seamless and powerful for companies around the world!

@zeiki_yu that was really our take with Hugo. Make it transparent

@ohansemmanuel Thanks!! Let's crush it

@ohansemmanuel Thank you a ton Ohans

@ohansemmanuel Thanks a mile! Feel free to spread the words about it all around you!

@samuelberthe 🤌

@maklyen_may TYSM for your comment and honest feedback! Are you keen on trying? It's free!

@maklyen_may It's a complementary product of Crisp, made by Crisp, but service a special purpose. Some users want an Human Inbox = Crisp, and others more AI Native solution = Hugo

@shubhagrawal26 If you know you know 👀 More than a positioning, fair pricing has always been at the core of Crisp since day one

@shubhagrawal26 😁

@mikestrives Thanks so much :)

@co_tunder Amazing! thank you for this cool feedback. And the cool thing is, it's already available to all Crisp and non Crisp customers. Ready to get started?

@jhumanj Thank you!! We're so happy to have Hugo in the Crisp family :)

@vincent_bradley thank you very much! If you ever want to stop paying 1$ per resolved tickets, you know where to go 😜

@quentin_decre Hugo as the version we release is just the beginning. We already have a solid roadmap and some heavy research on Hugo. Mostly auto-training systems, improvements on Copilot system as well. Goal of Hugo is to be Cursor / Claude Code like for customer support.

@milukove Thank you very much Egor

@panphilov Amazing! I'm curious, what makes you want to try Hugo?

@busmark_w_nika this one was not easy to be honest, because we had to rebuild our entire stack from scratch. Thank you for supporting us and hopefully Hugo will help you improve your customer support experience

@busmark_w_nika Thank you Nika :)

@yannick_mthy Thank you so much!

@vedranrasic Thank you so much!

@zongze_x Brilliant. Congrats on the launch! How do you prevent users from spending time on ideas that won’t sell?

@zongze_x Congrats! 🚀 looks like a serious step up from just coding. how does the research agent handle niche b2b markets compared to generic consumer data?

@cruise_chen 
Love that comparison. Race Mode is our way to make trade-offs explicit instead of locking you into one path. You get multiple approaches in parallel, then you pick based on criteria like speed to ship, complexity, and growth potential. Would love feedback on how we should present the comparison to make the choice even easier.

@cruise_chen Haha, exactly 😄 Race Mode is basically controlled competition without the politics. What surprised us most is where the teams diverge, not just in execution speed, but in assumptions about users, pricing, and distribution. Curious: if you could apply “Race Mode” to a real human team, would you optimize for speed, quality, or contrarian ideas

@cruise_chen Thank you, I love that comparison. That’s exactly the spirit: parallel options, less politics, and clearer choices.

@kate_sleeman 

Exactly. Early-stage building is usually fragmented across docs, chats, repos, and tools, and the original insights get lost. We’re trying to keep research, decisions, and execution in one continuous thread so the build stays aligned and iteration gets faster. Thanks for calling that out.

@kate_sleeman Thank you, this is exactly the problem we’re trying to solve. Keeping research, decisions, and execution connected is where the workflow really starts to feel “compound.”

@kate_sleeman Totally agree, and thank you for sharing that. Keeping research and execution connected is one of the main reasons we built Atoms this way.

@lvyanghuang 
Great question. Today the sweet spot is builders who want something shippable quickly, especially:
micro SaaS and paid utilities, niche tools, AI wrappers with real distribution plans, and long tail products where research and SEO matter.

DTC can work too, but we’re strongest when the core “product” is software and the loop is research → build → launch → iterate.

@lvyanghuang Thank you! Right now the sweet spot is indie SaaS and small paid tools, especially niche workflows where research and distribution matter and you want to get to something shippable with auth and payments. DTC can be a fit too, but we’re strongest when the core product is software.

@lvyanghuang Thanks. Right now the sweet spot is ideas where research, scoping, build, and go to market need to stay tightly connected. Think indie SaaS and small tools with a clear niche and a reachable distribution channel. DTC can work too, but it is usually more asset and brand heavy, so we see better early wins in focused B2B or prosumer workflows.

@libin_yao 
Thank you. That’s exactly the bet: not just “AI helps you code”, but an AI team that can run the full loop from idea to something you can actually ship and monetize.

We’re actively compiling end to end case studies now and we’ll share them publicly soon, including what worked and what broke. If you have a risky idea, drop a one liner here and we’ll suggest a fast MVP scope to validate it.

@libin_yao Thanks so much. We’re working on real end to end case studies now and we’ll share more soon. And risky ideas are honestly a great fit, because the goal is to reduce the cost of testing and learning quickly.

@libin_yao Thank you, really appreciate it. We are aligned on case studies being the proof, and we are working on sharing more real examples. If you try it on a risky idea, I would love to hear what felt most uncertain and whether the agents made those assumptions explicit.

@vega_chan 
Totally agree, and thank you. We’ve seen too many “instant build” tools push you into shipping something before you’ve sanity-checked the market.

Atoms tries to earn the right to build by doing research first: clarifying the target user and pain, checking alternatives, identifying a realistic distribution path, and then proposing a tight MVP. And if the idea looks weak,
we’d rather say that early and suggest a better angle than generate a shiny prototype that goes nowhere.

@vega_chan Thank you, that means a lot. We’ve seen too many tools rush to UI and code and skip the hard question of “should you build this at all.” Our goal is to help you sanity-check the market and distribution early, then scope a tight MVP based on that.

@vega_chan Thank you, that means a lot. We agree the best outcome is not more UI faster, it is knowing when the idea is weak early, with clear reasons and assumptions. We try to make the research agent surface red flags, unknowns, and what would need to be true before you invest more time.

@yehan_xiao 
Thank you. That connection is exactly the bet. We kept seeing teams generate prototypes quickly, then get stuck on the unglamorous parts: decisions, integration, launch readiness, and distribution. Atoms is our attempt to make that end to end loop more repeatable.

@yehan_xiao Thank you, I really appreciate that. Connecting those pieces is the hard part, but it’s also the part that matters most for solo builders.

@yehan_xiao Thank you. We know it is ambitious, and we appreciate the respect. Our goal is exactly what you said, connect research, build, and go to market into one coherent loop.

@busmark_w_nika 
Great question. Tools like v0 are amazing for fast UI and prototyping.

Atoms is built for the “idea to business” loop: it starts with research and product scoping, then carries that context through build, backend essentials like auth and payments, and finally launch and distribution (SEO/growth). So the goal isn’t just a nice UI, but something closer to shippable and monetizable.

@busmark_w_nika Great question. v0-style tools are fantastic for fast UI and prototyping. Atoms is built for the full idea to business loop: research and scoping first, then build with the “messy middle” like auth, payments, deployment, and a distribution plan so it’s closer to something you can ship and charge for.

@busmark_w_nika Great question. Tools like V0 are awesome for generating UI and code quickly. Atoms is aiming to be a full AI business team workflow, research, positioning, product spec, build, then go to market outputs like landing copy and distribution plans, with the goal of turning an idea into something you can actually ship and charge for. If you tell me what you use V0 for today, I can map the overlap and the differences more concretely.

@madalina_barbu Thank you so much! Really appreciate the support.

@madalina_barbu Thank you. Really appreciate it.

@madalina_barbu Thank you so much.

@gxy5202 

Really appreciate that. We agree the space is early and there’s tons of room to improve. Our focus is reliability and completeness: stronger research before building, clearer trade-offs, and a workflow that gets you beyond “AI-generated output” to something you can actually ship and iterate. If you try it, I’d love to hear where it feels better and where it still falls short.

@gxy5202 Thank you! Totally agree the space is still early and there’s a lot of room to improve. We’re putting most of our energy into making the workflow more reliable end to end, not just generating something that looks good. Really appreciate the encouragement.

@gxy5202 Thank you so much. Really appreciate the thoughtful encouragement, we are definitely trying to push the quality bar beyond “just generate something.”

@valeriia_kuna 
Great question, and this trade-off is exactly where most “autonomous” systems go wrong.

In Atoms we try to handle it in a few explicit ways:

• Separate “ship fast” choices from “hard to undo” choices
We’ll move quickly on reversible work, but we surface high-impact decisions (positioning, ICP, pricing model, acquisition channel focus, data model, auth/payments approach) for your approval instead of silently optimizing for speed.

• Make the trade-off visible, not implicit
When the team proposes a faster path that could hurt long-term GTM, we present it as options with pros and cons, and we ask you to pick the priority (time-to-ship vs GTM leverage vs maintainability).

• Keep a GTM constraint in the plan
If your goal is SEO, PLG, or a specific channel, we treat that as a constraint upfront so build decisions align with distribution, not just engineering convenience.

• Use parallel proposals when it’s ambiguous
With Race Mode, we can explore a “fast MVP” path and a “GTM-aligned” path in parallel, then compare on criteria like time-to-first-signal, time-to-first-dollar, and long-term channel fit.

@valeriia_kuna Thank you, and great question. When we see a “faster to build” option that could weaken long-term GTM, we try to make that trade-off explicit instead of hidden. We surface options with the key assumption behind each, and we ask you to pick the priority for this iteration, for example fastest signal vs channel fit vs maintainability. For irreversible choices, we add a checkpoint so you can approve before the team commits.

@valeriia_kuna Thanks, and yes this trade off is the core. We make the team propose options with the “why” behind them, what you gain now, what you pay later, plus the key assumption. Then you can choose to bias toward speed, or toward long term GTM and moat, rather than the system silently picking the fastest build.

@candyrorae 
Absolutely agree, and thank you for putting it so clearly. That “setup cost” is the real tax that kills long-tail ideas.

Atoms is built to compress that cost by running the full loop as a team: research the niche, scope a tight MVP, ship with the boring essentials (auth, payments, deploy), then iterate based on real signals. The goal is exactly what you said: make it economical to test many small niches and double down on the ones that show traction.

If you have a niche in mind, drop it here. Happy to suggest what a “10 in a day” style MVP would look like for it.

@candyrorae Yes, exactly. Setup cost is the killer for tiny niches. If we can make “research → ship → test distribution” cheap and fast, then you can run many small bets and only double down when you see real signal. Thanks for articulating it so well.

@candyrorae This is exactly the bet that excites us too. When the setup cost drops close to zero, tiny niches stop being “not worth it” and start being a real strategy. We are building Atoms around that long tail economics shift.

@shirleyw 
Yes, it’s on our roadmap. Right now Atoms is optimized for solo builders and small teams, but we’re actively working toward collaboration features like shared workspaces, roles and permissions, project history, and handoffs so multiple people can co-own a project cleanly. If you tell me your team setup, I’d love to learn what collaboration workflow you need most. 

@shirleyw Yes, it’s on our roadmap. Right now Atoms is optimized for solo builders, but we’re actively planning collaboration features like shared workspaces, roles and permissions, project history, and clean handoffs so multiple people can co-own a project.

@shirleyw Yes. Collaboration is on our roadmap. Multi person projects with shared context, roles and permissions, comments, and clear version history are all important for this to work in real teams. If you tell me your typical setup, solo plus occasional collaborator, or a full team, I can share what we are prioritizing first.

@31xira 
Really appreciate this, and you described the problem perfectly. The “messy middle” is where most projects stall.

Atoms is designed to make those parts more repeatable: we keep research and product decisions attached to the build, surface trade-offs for approval instead of silently deciding, and try to automate the boring but critical wiring like auth, billing, and deployment with production-friendly defaults. We’re still iterating hard on consistency, so if you try it, your feedback on where it breaks or feels uncertain would be incredibly valuable.

@31xira Thank you, and you nailed the pain. The messy middle is where most prototypes die. We try to make those decisions and trade-offs explicit, keep research and scoping connected to implementation, and use production-friendly defaults for things like auth, billing, and deployment so it’s easier to get to something stable. Consistency is a big focus for us right now, so if you try it, your feedback on where it feels unclear would be incredibly valuable.

@31xira Thank you, this means a lot. The “messy middle” is exactly what we want to solve: making decisions explicit, showing trade offs, and handling the unglamorous wiring like auth, payments, deployment, and go to market steps so solo builders can ship something real.

@jeetendra_kumar2 
Great question. Many existing tools are excellent at generating UI or getting you to a prototype quickly. Atoms is designed around the full “idea to business” loop, so it starts earlier and goes further.

What’s different in practice is:

• we begin with market research and product scoping so the build is grounded in a real user and distribution angle

• we carry that context through implementation, including backend essentials like auth, data, and payments so it’s closer to shippable

• we don’t stop at code generation, we also help with launch and iteration loops (SEO and growth planning)

So the focus is less “build faster” and more “ship something people can actually use and pay for.”

@jeetendra_kumar2 Great question. Many coding tools are excellent at getting you a prototype quickly. Atoms is built around the full “idea to business” loop: research and scoping first, then building with the unglamorous production pieces in mind like auth, payments, deployment, and a distribution plan. The goal is not just a working demo, but something you can actually ship and charge for.

@jeetendra_kumar2 Great question. Most coding tools optimize for getting code and UI fast. Atoms is built around the full path from idea to a chargeable, operable product, research, positioning, spec, build, and go to market, with decisions tracked as assumptions instead of just generating screens.

@zeiki_yu Thank you. That end to end loop is exactly what we’re building for: helping solo founders go from idea to something live, measurable, and monetizable, not just a prototype. If you try it, I’d love to hear what kind of business you build first.

@zeiki_yu Thank you so much. Solo founders were exactly who we had in mind when we built the end to end workflow.

@zeiki_yu Thank you. Solo founders are a big focus for us, keeping the loop tight from idea to something live.

@villazhao 
Thank you, that means a lot. We’re obsessed with the unglamorous “make it real” pieces, because that’s where most prototypes die. If you have an idea you want to monetize, share a one-liner and we’ll suggest a fast MVP scope to get to first dollars quickly.

@villazhao Thank you, that means a lot. We’re pretty obsessed with the “make it real” parts that usually stall solo builders. Rooting for you back, and if you try it, I’d love to hear where it feels smooth vs where it still needs polish.

@villazhao Really appreciate it. We are rooting for the same thing, less shiny demo, more real shipping with the boring but necessary pieces.

@qiwap 
Love that framing, thank you. That’s exactly how we think about it too: the value is in keeping research, decisions, building, launch, and iteration connected so context doesn’t get lost. If you try it, I’d love to hear which step in the chain feels strongest and which step still needs work.

@qiwap Thank you, I love that phrasing. That’s exactly our belief too: the value is in keeping the whole chain connected so context doesn’t get lost between research, decisions, building, and go-to-market.

@qiwap Thank you. That is exactly how we think about it too, the workflow and the handoffs are the product.

@daxin_wang 
Totally agree, stitching these workflows together is the hard part.

Our approach is to treat this as a coordinated team, not a single model trying to do everything:

• specialist agents own each domain (research, product, engineering, SEO, analytics), with a Team Lead agent keeping the overall goal and constraints consistent

• each domain runs with structured checklists and outputs (for example research generates positioning and target queries, SEO turns that into an information architecture, content maps back to distribution intent)

• we keep a shared project memory so later steps reuse earlier assumptions instead of re-inventing them

• we validate with “execution reality” so content and SEO plans stay aligned with what we can actually ship and measure

@daxin_wang Totally agree, that stitching is the hard part. We manage it by treating Atoms like a coordinated team: specialist agents for research, product, engineering, SEO, and analytics, plus a lead agent that keeps a single source of truth for assumptions and decisions. The key is making the handoffs explicit so SEO and content stay tied to the actual product, ICP, and distribution strategy.

@daxin_wang You are right, it is hard. Our approach is to use structured playbooks per domain, require sources and citations in research, and keep SEO and content tied to the same ICP, positioning, and keyword intent so they do not drift. We also surface uncertainty and “needs human judgment” flags when the system cannot reliably know.

@frank_li13 
Love that, thank you. That’s exactly what we mean by “vibe-business”: not just turning creativity into a prototype, but pushing it all the way to something you can actually launch, charge for, and iterate on.

And I agree on the bigger picture too, it changes how new businesses can get funded and started. If the cost of testing an idea drops dramatically, more creators can bootstrap and validate without needing capital upfront.

If you share the kind of creative business ideas you’re thinking about, I’m happy to suggest what a fast “first revenue” version would look like with Atoms.

@frank_li13 Thank you, I really love how you put that.

@frank_li13 Appreciate that. “Vibe business” is our way to describe the gap between having a cool idea and turning it into something that can actually earn revenue. If it helps more creators cross that gap, we are doing our job.

@sam_chen1 
Great question. Today we help most with the planning and production side: identifying target audiences and angles, generating positioning and content ideas, drafting posts, and building a simple content calendar aligned with your product and distribution strategy.

We’re not trying to “spam autopost” everywhere, because quality and authenticity matter. The goal is to help solo founders move faster while staying on brand and staying human.

@sam_chen1 Totally agree. Today we help most with the strategy and execution support: defining audience and angles, writing channel-specific drafts, building a lightweight content calendar, and keeping it consistent with your positioning. We’re not trying to be a spam autoposter. The goal is to help solo builders ship authentic, on-brand marketing faster.

@sam_chen1 Thanks. For social, we help you go from ICP and positioning to a content plan, post angles, drafts, and a lightweight distribution checklist. We do not pretend there is a magic autopilot, but we try to make the work clear, repeatable, and tied to the same strategy as the product.

@anthony_cai 
Yes, that’s the direction, and we try to cover SEO as a system, not just “write a blog post.”

Atoms can help with:

• topic and keyword research plus content outlines

• generating SEO oriented pages (programmatic or editorial style, depending on the product)

• on page basics like titles, headings, and metadata suggestions

• internal linking structure recommendations so pages support each other

How “automatic” it is depends on your project and how much control you want. If you tell me your use case (content site, directory, micro SaaS, tool pages, etc.), I can share the exact SEO workflow we recommend and what gets generated today.

@anthony_cai Yes, that’s the direction. We help with topic and keyword research, outlines, SEO-focused pages, basic metadata suggestions, and internal linking structure so it’s not just isolated content. How automatic it is depends on your project and how much control you want, but we aim to treat SEO as a system.

@anthony_cai Yes, we can help with SEO outlines and briefs, draft pages, suggested internal linking structure, and metadata like titles and descriptions. We also try to tie it back to a keyword and intent map so it is not just generic content.

@osakasaul Thank you for spending the time, and for doing the survey. We really appreciate it.

@osakasaul Thank you for spending the time, and for doing the survey too, we really appreciate it. We’re reviewing submissions in batches. If you can reply with your PH username, I’ll help make sure we locate your survey and get you an update.

@osakasaul Thank you for building with Atoms for several days and for filling out the survey, that really helps. We review submissions in batches, so it can take a bit. If you do not hear back soon, please DM me your signup email or a screenshot of the confirmation and we will check it for you.

@dessignnet Thank you. If you try it, I’d love to hear what idea you test first and where the workflow feels smooth vs where it still needs polish. Launch-day feedback is incredibly valuable for us.

@dessignnet Thank you. Would love to hear what you try building first and what feels smooth vs confusing.

@dessignnet Thank you. Would love to hear what you try first and what feels unclear once you test it.

@jaceperry  @tripplep 
This made our day, thank you. We’ve put a lot of effort into keeping the exported code maintainable with clear separation of concerns, because we want people to be able to own and extend it.

And yes on “vibe marketing”, we love that direction. Today we already support SEO and distribution planning, and we’re exploring deeper marketing workflows once a product is shipped. If you have a preferred channel (X, LinkedIn, SEO, outbound), I’d love to hear what “vibe marketing” would look like for you.

@jaceperry  @tripplep This made my day, thank you. We care a lot about exported code being maintainable and easy to take over, so I’m really glad the structure felt solid. And yes, vibe marketing is such a good idea, we’re excited about building more “after shipping” workflows. If you have a preferred channel, I’d love to hear what you’d want that feature to do first.

@jaceperry  @tripplep This is awesome to hear, thank you for trying it and for the concrete feedback on code quality. And +1 on vibe marketing, we also think distribution needs to be part of the same loop, so it is definitely on our radar.

@kilpatrick 
Thank you so much. Great question, and it’s the right bar to hold us to.

We’re actively compiling a set of real end to end case studies with links and clear breakdowns of what Atoms did vs what the human decided, plus what worked and what didn’t. Some early projects are public already, and a few are still in private beta while teams finish polishing.

@kilpatrick Thank you so much. Great question, and it’s the right thing to ask. We’re putting together a set of real end to end case studies with live links and a clear breakdown of what Atoms did vs what the human decided. Some early projects are already public, and a few are still polishing before they open up to new users. If you share what type of business you’re most interested in, I’ll point you to the closest examples we can share right now.

@kilpatrick Thank you so much. We are still early, but yes, we are collecting and sharing real shipped examples and case studies as they go live. If you tell me what kind of business you want to see (B2B, prosumer, DTC), I can point you to the closest examples we publish next.

@mrpop 
Yes. You can use Atoms purely for research and evaluation: market and competitor analysis, positioning, ICP clarity, distribution hypotheses, and MVP scoping, without committing to building or launching. A lot of users start there to decide what’s worth pursuing before spending time and budget. If you share a rough idea, I’m happy to show what the evaluation output looks like.

@mrpop Thank you! We put a lot of effort into keeping it clean, because the underlying workflow can get complex quickly.

@mrpop Thanks, and yes we are thinking in that direction. Once something is shipped, the next unlock is consistent distribution. We want marketing to be part of the same loop, not a separate set of tools, so the product and the growth plan stay aligned. 

@jaceperry 
Yes, we’re very interested in that direction. Shipping is only half the battle, distribution is where most products stall.

Today we already help with distribution planning, SEO workflows, and content creation. “Vibe marketing” as a deeper feature set is on our roadmap, things like channel specific messaging, content calendars, landing page testing, and iteration loops tied to metrics, without turning it into spam automation.

Also, we will public the Ads Agent soon

@jaceperry Yes, we’re very interested in that direction. Shipping is only half the battle.

@jaceperry Thank you. Research and making the bottlenecks visible is a big focus for us, and we are pushing hard on clarity, not just output volume. And yes, the “build and launch” part is ambitious, we are excited to keep iterating with builders like you.

@peterbuch Congrats! Do you monitor for AI hallucinations in generated content?

@peterbuch best of luck mate. looks excellent!

@zeiki_yu thank you for the support, let me know if you have any questions on findable

@zeiki_yu thank you for your kind words, please let us know what you think of the platform

@alina_petrova3 great question!

Right now the findable score is based on what models know about the brand without access to search, robots.txt settings for crawlers, web performance metrics and a trust model (modeling Google's EEAT guidelines).

That said, we are adding more signals to the findable score.

The idea is to provide a reasonable score to benchmark against + ability to improve systematically.

@flundberg_at_incredible great feedback. We are working on adding even more export, reporting and collaboration features. Stay tuned!

@jeetendra_kumar2 Hi Jeetendra, can you elaborate what you mean? tahnk you

@jeetendra_kumar2 you can just sign up and set up your brand and competitors in findable (takes a few seconds). From that point on findable will help you understand what ChatGPT, Grok, Google AI Mode, Gemini, Claude et al think about your brand and your competitors, why and what you can do about it.

@harkirat_singh3777 thank you, Harkirat.

@peterbuch  @harkirat_singh3777 can't agree more. Super happy to be able to roll out the free plan so everyone can get started with findable as full featured Answer Engine Optimization platform.

@german_merlo1 thank you, Germán. Let me know what you think?

@german_merlo1 thank's a lot for checking out findable. Looking forward to your feedback!

@ahmer_saud thanks for trying, please let me know how we can make findable even better for you

@ahmer_saud thanks a lot for giving findable a try.

I'm especially interested in what you think about the "Content" tab where we run a full content gap report for you.

Let us know if you find anything we can make better.

@jorgealonsodf thank you, Jorge, I appreciate the support.
If you are optimizing your page for SEO it certainly helps. But AEO goes beyond what's on your website. With findable we show you how LLMs like ChatGPT see your brand, and how they talk about you. Also what you can do to rank better. So SEO is 70% of the work, but the real growth opporuntities are often in the sources that LLMs use.
Happy to jump on a call to chat more.

@peterbuch  @jorgealonsodf like @peterbuch already said: SEO definitely helps and in a sense AEO is an extended version of SEO (and arguably modern SEO is AEO) so it is a bit tricky to answer the question without getting into word definitions :D

I think the good news is: if you have an existing, strong SEO foundation you do have a head-start when it comes to AEO.

But the flip-side is also true: if you start from scratch, with a new brand/startup/product and even if you do not have a strong SEO background there are tons of things you can do to get more visibility.

AEO is so broad and there are so many angles to improve (on-site content, off-site content, social, press, training material, video, search, …)

That's why I'm so excited about this launch: you can just sign up for free, set up your brand and immediately benefit from findable and you get continuous monitoring, reporting and content opportunities our of the box.

@juannikin thank you, appreciate your support. You really are everywhere :)

@gokuljd great question. In some cases you can improve your visibility almost instantly.

Seconds to minutes (!)

Some of our customers find actionable low hanging fruit as soon as the first findable reports are done (a few seconds after signup).

Why is this possible so fast?

Because ChatGPT and co use search heavily.

It depends on the space, personas, topics, prompts, competition and so on of course.

But to give you a concrete example: launching on Product Hunt is an excellent way to get onto the map fast. Especially when your product or service is still a rather unknown.

Why? Product Hunt ranks well on SEO so it is quite findable by search tool use of ChatGPT, Gemini, Claude and so on.

But more than that: the models know that Product Hunt and the community around Product Hunt (similar to reddit) are a destination / directory that gives orientation.

Also there are a lot of forums, blogs and newsletters that talk about what's new on Product Hunt, which again helps with SEO but also with visibility in model training data long term.

So while it might take a while to get SEO on your own website going you can make progress fast when you get into the right directories or other off-site destinations like reddit, podcasts, magazines, social media and so on.

That said: don't underestimate longer-term efforts like investing in owned media + your own SEO.

Please let us know how findable works for you and if you found low hanging as well as longer term stuff that helped you.

@chrismessina Congrats on the launch Chris! How do users interact with multiple agents? Through a single interface, scripts, or dashboards?

@alexmturnbull This is genuinely impressive.

A few things that really stand out:

• 65% resolution guarantee is bold; and rare. Most “AI support” tools avoid committing to outcomes.

• Actually closing tickets end-to-end (billing, plan changes, invoices) instead of drafting replies is the real unlock.

• The VIP concierge + hands-on setup shows you care about results, not just demos.

• From a user POV, this removes friction instead of adding another layer of automation.

This feels meaningfully different from typical AI copilots (and frankly far more useful than tools like Intercom Fin that still leave humans doing the work).

Congrats on the launch; this solves a real problem the right way.

@alexmturnbull Brilliant! Congrats on the launch Alex. How do you handle multi-agent collaboration or handoffs?

@alexmturnbull 🙌

@benln  @tmorkes lets gooooooooooo!

@rosekamalloveGood question. The toughest ones were edge cases where even humans didn’t follow a consistent process. Fix the inputs, and Helply gets reliable fast.

@rosekamallove Thanks! The hardest ones are very often the ones that seem easy, but require a context that is not provided in the Knowledge Base articles. Thanks to the features like Gap Finder and Guidance, our AI Agent can easily tackle those questions reliably. As an example, we have a client that use mentioned features to answer questions for legacy and new tools they provide. Their AI Agent can easily understand what tool is being used by the way the question is asked. If it's still missing that context, it asks a follow-up questions, ensuring very high level of accuracy

@aavishkarmishra thanks for the comment! Love the feedback. Happy to show you how it works and get you set up with a custom AI agent free to test it out and prove to you what we can do :)

@aavishkarmishra Thanks Aavishkar, appreciate that. Removing tickets entirely was the bar we set from day one.

@alexmturnbull  @brent_hardinge it's great working with you Brent. Thank you for sharing your stats so transparently. Excited to continue working with Adventist into 2026 and beyond!

@brent_hardinge thanks Brent, really appreciate you sharing the results. Excited to see it expand across the team :)

@alexmturnbull  @brent_hardinge inject this directly into my veins :) Love to hear these numbers.

@why_tahir Great question. This is exactly the part most AI support tools gloss over.

A few core principles we built Helply around:

1. Confidence thresholds, not blind automation
Helply only takes real actions (billing, plan changes, account updates) when it’s highly confident. If confidence drops below a defined threshold, it escalates immediately with full context instead of guessing.

2. Explicit permissions by action type
Teams control what Helply is allowed to do. You can enable read-only, suggest-only, or execute actions per category (billing, account, refunds, etc.). Nothing is on by default.

3. Guardrails + reversible actions
Actions are constrained, logged, and reversible. No open-ended “do anything” access. Every step is auditable.

4. Human-in-the-loop where it matters
For sensitive workflows, Helply can propose the action and wait for approval. Over time, teams can loosen that as trust builds.

5. Outcome-based enforcement
We back this with a 65% resolution guarantee. If Helply isn’t behaving safely or accurately, that’s on us, not the support team.

Net-net: Helply escalates early rather than risk doing the wrong thing. Trust comes from restraint, not bravado.

@why_tahir love it. great question.

@marioandrearaujo that's awesome. Can't wait to have you test it out and get your feedback! Let's go! 🚀

@marioandrearaujo thanks Mario! 20 years?! I'm feeling old LOL. Thanks for following all these years and excited for you to give Helply a spin when you're ready :)

thanks @zeiki_yu! we are so proud of this product and appreciate you checking it out :)

@alexmturnbull  @jared_scheel  @harkirat_singh3777 thanks for the love Karkirat! Let's go!! 🚀

@tmorkes  @jared_scheel  @harkirat_singh3777 Thanks Harkirat. Early adoption usually comes down to two things:

1. Where the AI agent is placed in the user journey

2. Ticket volume

Teams with higher volume that surface the agent in a high-intent spot (help center, contact flow, billing pages) tend to see adoption almost immediately. That’s also why we’re comfortable backing the outcome with a guarantee.

thanks @dylan_hey! Let's go!! 🚀

@dylan_hey thanks man! We'll be sharing it all :)

@tmorkes  @utsavpm thanks man!

@leonie_fischer We only allow real actions when intent is explicit, confidence is high, and permissions are enabled for that action. Nothing is implicit or on by default.

Every action is scoped, logged, and reversible. If confidence drops or intent is ambiguous, Helply escalates instead of acting.

@leonie_fischer We always suggest to start with "read only" actions (getting the last invoice, subscription status etc.). When our clients are gaining confidence in our agent, they tend to roll more actions in a time. Apart of our internal safeguards, users can also specify when action should run.

@shreya_chaurasia19 Good question. We don’t think in terms of “AI vs human” but risk, reversibility, and confidence. Low-risk, easy-to-undo workflows can run end-to-end. Anything with financial impact or a big blast radius requires confirmation or a human check.

Teams start conservative and loosen controls as Helply proves itself on their data. If confidence drops, it escalates early with full context.

@shreya_chaurasia19 we've found that highly-complex issues tend to require escalation to a human, regardless of what agent you are using. Way too many people were over-eager about dumping everything to AI (see Salesforce admitting that they over-promised). We did a lot of work around making sure Helply can answer questions but also bow out of the conversation when it should. Ideally that's the exception and not the rule, though! In the future we plan to explore having more support for mid-conversation long waiting human-in-the-loop checks for when escalation isn't the answer but you still someone to push the button.

@alexmturnbull  @dosberg I love this Doug. Thanks for sharing your story. Alex was a huge inspiration for so many (myself included!), and it's been nothing short of amazing working with him on this. You will definitely see the love and focus we've put into Helply. Appreciate you :D

@dosberg thanks Doug! That blog changed my life and I'm grateful everyday I was able to get lucky with it back then. Cheers to many more years building in public. Here's an image I dug up from that first post haha

@saurabh_ishere woohoo! Let's go!!! 🚀

@saurabh_ishere Thanks! Totally agree. Most tools just move work around. The hardest workflows so far haven’t been billing mechanics but intent-heavy edge cases. Things like refunds or plan changes where the user is frustrated, unclear, or mixing multiple requests in one message.

The automation itself is solvable. Getting intent right with zero blast radius is the real work, which is why Helply escalates early when signals aren’t clean.

@shivali_goyal1 hey there :) Curious how you handle accuracy and edge cases: when a screen recording includes detours, mistakes, or non-linear flows, how does AI Doc Writer decide what becomes part of the final step-by-step doc? Is there a confidence/scoring or review layer to avoid hallucinated or misleading steps? This is a really interesting shift from better editors to no writing at all!

This is brilliant! Automating documentation from screen recordings solves a real, painful problem. The "light years ahead" feedback says it all. Excited to try the free trial and see it in action. Thanks for sharing!

@shivali_goyal1 Congrats on the launch! How does AI Doc Writer integrate with existing workflows—Word, Google Docs, email?

@abod_rehman so true! That's what we kept hearing from everyone.
Do try it out and let us know your experience

@abod_rehman Totally agree.
That’s exactly what pushed us to build this.

@abod_rehman that's why we have created this ;)

@a_s_k_af thanks!

Thanks Arun, really appreciate the kind words 🙌
We’ve put a lot of care into making the output feel genuinely “shippable”, not generic.

If you try it, I’d love to know what you’re documenting most often: internal SOPs, product walkthroughs, or customer onboarding?

@a_s_k_af thanks man! Happy to see Trupeer being helpful in your work

Thanks @harkirat_singh3777 
That “messy recordings → clean, on-brand docs” line is exactly the goal.

We built Trupeer to work with real-world inputs (Zoom/Meet demos, imperfect walkthroughs), then use vision-based understanding to pick the actions that matter and turn them into a doc that reads like something your team would actually ship.

Curious: where would this help you most, internal SOPs, customer onboarding, or product docs?

@shivali_goyal1  @harkirat_singh3777 thanks buddy! It's so good to see you here

@prashastha_jain you bet! ;)

@prashastha_jain Thanks, really appreciate it 🙌
We’re trying to make docs feel as easy as hitting record.

@prashastha_jain you bet ;)

@pritish_gupta2  @huisong_li thanks!

@huisong_li thanks!

@huisong_li Really appreciate your support.

@iamarnob6543 
This means a lot, thank you ❤️

We’ve been obsessed with the nitty-gritty because that’s where documentation usually breaks. People don’t struggle to “explain” something; they struggle to turn that explanation into a doc that’s actually usable and shippable: clean steps, correct screenshots, and the same structure your team expects, without spending hours formatting and rewriting.

That’s why we built Trupeer to work from real recordings (including messy Zoom/Meet demos), use vision-based understanding to detect meaningful actions, and then generate a doc that reads like something a team would confidently send to customers or ship internally.

If you get a chance to try it, I’d love to know what your primary use case is at Olvy: release notes, product docs, onboarding, SOPs, or internal workflows. We’re actively iterating based on feedback, and your perspective would be super valuable

@iamarnob6543 thanks a lot for all your support! It's great to see you here again

@iamarnob6543 thanks a lot buddy!

@kritikasinghania Thanks 🙌, really appreciate it!

We built this because most teams already explain processes on recordings, the painful part is turning that into a clean, on-brand doc that’s actually usable. Trupeer does that in minutes from screen recordings or even Zoom/Meet demos.

@chilarai really appreciate it!

If you get a minute to check it out, Trupeer turns everyday recordings (walkthroughs, demos, Zoom/Meet calls) into clean, on-brand docs in minutes. Would love to know what you document most often at Swytchcode: SOPs, product guides, or training docs?

@chilarai thanks!

@shubham_pratap appreciate your support.

@sam_chen1 great to know! Do you currently use any doc product currently?

@jebank wooow, love this!!

@jebank Congrats on the launch Jacob! How do the agents coordinate tasks and avoid conflicting actions?

@jebank congrats on the launch. Is it possible to easily orchestrate multiple agents?

@dougli love that you left OpenAI to take on this massive challenge and super impressed by what you have already built!

@dougli Congrats on the launch! What’s the first moment users realize it’s truly ‘telepathic’?

@dougli Congrats on the Gmail rollout! Curious to see how users adapt to writing with full context like this. Definitely feels like a big shift in daily communication. Also curious, what’s been the most surprising feedback since the Gmail launch?

@chrismessina We're delighted to have you as our Hunter!

One clarification on data processing -- we do process remotely (big GPUs needed to power the models), but we never store your data or train on it without consent.

Here's to telepathic communication :)

@chrismessina fully agree! been using for the past 6 weeks and it is insanely good

@chrismessina finally an ai tool in the email space that isn't just a gimmick but delivers real impact

@luissa_schemuth Thanks Luissa! You're right - I never would have thought replying to emails could be this addictive!

@luissa_schemuth Totally agreed!!

@luissa_schemuth I second this – Luissa is spot on. The way it integrates directly into Gmail is a total game-changer and it keeps my workflow so clean. Huge kudos to Doug, Matt, and Marcus for making the inbox experience so seamless!

@osmanio2  Thanks for the kind words!

@marie_brayer_ftw 

"It is so good that I'm starting to use Gmail as a lightweight notetaker" hahaha I do this too! We're working on getting this working in other places, can't wait to release new features for you to try.

Thank you for being such an early tester! We couldn't have done this without you :)

@jesperht Thanks Jesper. Always appreciate the support

@kevan_williams Thanks Kevan! Really appreciate the feedback.

@kevan_williams Thank you - really appreciate that. Authenticity is exactly the hard part we cared most about. We’re trying to build something that supports how people already communicate, rather than replacing it with a “better” but generic version. Feedback like this genuinely means a lot to us 🙏

@valeriia_kuna Thanks Anna for the kind words! I agree - personalisation is important.

We have mechanisms to balance past thread context with the user’s global writing style. Both approaches are valid; what matters is having the most accurate, up-to-date version of the user’s style and feeding it to the AI at the right time. Context engineering is central to our product, and we’re focused on making it correct and reliable for our users.

Can't wait to see what you've been working on 👀 Fingers crossed for your product!

@zeiki_yu Thank you Zeiki! Glad that this resonates with you!

Thanks for the feedback!

@mleelightfern  @dougli  @harkirat_singh3777 Thanks for the kind words 🙇‍♂️

@linh_s Thanks for helping us test out Lightfern! Really appreciate all the feedback

@byalexai My first output: https://grok.com/imagine/post/025c2978-d993-4c5c-af42-9a1aebf2ffec?source=post-page&platform=web :D

@byalexai Congrats on the launch Alex! Brilliant for creatives and a fun for turning ideas into reality. Does Grok Imagine generate outputs fully automatically, or guide users step by step? How does the AI handle vague or abstract ideas?

@yahia_bakour3 Great product Yahia! Congrats on the launch. How do you manage permissions or trademark issues?

@zeiki_yu Thanks!

@ltatis thank you!

@polnikale appreciate you as always man

@tomaslau Thank you, great to see you here as well :D

@valeriia_kuna The algorithm's pretty simple at this point. We track the topics you explore in each session and how much you engage with them, then use AI to suggest new topics you'd likely find interesting!

@victor_eth Wikipedia was a metaphor haha -- we actually don't use wikipedia to generate content, for now. But, using wikipedia links in a page to generate related topics would be an interesting idea to explore in the future. The onboarding still needs a lot of polishing but let me know how it goes! :)

@zeiki_yu Thanks Zeiki. Means a lot.
Any features you would have us add to it?

@victor_eth Hey Victor, share the sentiment you expressed. It is frustrating to keep reiterating yourself.

On non-OpenClaw stuff: all sync is powered through the chrome-extension. It periodically also syncs non-desktop sessions (if you grant permission). For OpenClaw: The agent invokes our APIs on interactions.

On privacy: we operate on zero-trust, but available-everywhere principle. So we store on a cloud-vault for access and performance reasons. All LLM calls are made in saving-off mode. All data is encrypted at rest. And air-gapped architecturally. Happy to answer any more questions.

Can't wait for you to try and share your feedback.

@ksenia_sh Thnaks Ksenia.

The OpenClaw plugin is agnostic of browser (integrates at API-level). The rest of the LLMs' context(Chatgpt, Claude, etc.) is currently supported through a Chrome-plugin (works on other Chromium-based browsers such as Brave as well). Which browser do you use?

@valeriia_kuna Nice idea on the memory-profiles. For now, it has to be 2 accounts. Would love to understand more on what all use-cases would need someone to have multiple memory profiles.

I've ended up juggling terminal output, browser tabs, and my editor on every agent run. Ray 3.0 turning that into a dedicated output window beside the code feels like a real win. The MCP server gets even better if it supports structured event streams and secret redaction by default.

@marek_nalikowski Thanks Marek!

@susan_pan1 Thank you, Susan, for loving our new feature! With Publish, we want to go one step further than just transcription or summaries, it turns meetings, audio, PDFs, and even videos into ready-to-share content like slides, blog drafts, or structured insights, so ideas don’t stop at consumption but actually get published and reused.

@ray_luan Thank you so much for the support! Let's keep going!!!

@eeeeeach That’s awesome to hear, thanks for sharing this! 🙌
YouTube → structured blog is one of our most-loved flows, especially for newsletters and Substack writers. We’re aiming to make sure the output feels publish-ready, not just a rough summary.
If you end up publishing one of those drafts, we’d love to see how it performs in the wild 🚀

@rocsheh Thank you so much, really appreciate the thoughtful feedback! 🙏
Custom export templates are actually something we’re actively exploring. Letting teams define their own structures (like decisions, action items, highlights) is a great way to make outputs consistent and reusable. Would love to learn what templates you find yourself repeating most in your work 😊

@vermouth2333 Yep, exactly 😄 Just drop a Web URL and HyNote turns it into a narrated summary you can listen to on the go. Perfect for long reads you don’t have time to sit through. Would love to know what kind of content you’re planning to try first!

@new_user___1282025165cc92287e7a197 This makes us so happy to read! Private podcast feeds are one of our favorite use cases too — it turns knowledge sharing into something people actually want to consume. Curious how your team listens: async updates, reviews, or learning sessions?

@jianqiang_hao Love that comparison, thank you!
That’s exactly where we see HyNote fitting in: not just understanding content, but helping creators package and share insights easily. Excited to see how you use the publishing side in your own workflow.

@mooyan That’s incredibly encouraging, thank you for following along! The “Publish” workflow was built to close the gap between thinking and sharing, so hearing this really validates the direction. If you end up publishing something with it, we’d love to see how you’re using it in practice!

@jayzhu Thanks so much, that means a lot! 🙏 Handling technical terms and domain-specific language was a big focus for us, especially for PDFs and research-heavy content. Glad it came through. If you throw even more complex material at it, we’d love to see how it performs for you.

@zephyrlink_i That’s awesome to hear, welcome aboard! 🚀 Really excited for you to hear your first conversion too. If you try turning it into a narrated summary or a publishable insight, let us know how it feels. Curious what type of content you’re converting first.

@justin_press Love this question, traceability is something we think about a lot.
Each polished output in HyNote keeps a clear link back to its original sources (meeting audio, PDFs, URLs, screenshots), so when summaries are reused or shared externally, you can always trace where an insight came from. Internally, this also helps teams iterate on outputs without losing context. We see this as essential if AI-generated insights are going to be trusted, not just consumed.

@klara_minarikova Great question, this is exactly the scenario we designed HyNote for.
Yes, HyNote automatically identifies action items, decisions, and next steps, and groups them into structured sections you can review or publish right away. In workshops, many teams export these as follow-ups or internal briefs so commitments don’t get lost after the session.
Would love to hear what kinds of outputs work best in your training flow!

@daniele_packard Thank you! 🙌 We’ve put a lot of care into making outputs flexible, because different teams and individuals need different formats. Hope you have fun playing around with it, would love to hear which export ends up being your favorite.

@daliuai Love that you called this out! We totally agree that capture is only step one. The auto-flashcards are our way of making sure insights actually stick, not just get archived. Curious: are you using them more for learning, review, or sharing with others?

@yuki1028 That means a lot, thank you! 🙏
“Active” is exactly what we’re aiming for. We don’t want notes to just sit there, but to be listenable, reusable, and easy to turn into real outputs. If you try turning one of your notes into slides or a draft, I’d love to hear how it fits into your workflow!

@onlyyixxs Thanks so much, Nathan! 🙌 The answer is YES, we already have mobile apps. You can capture ideas, review notes, and manage your Second Brain anytime, anywhere.

We’re continuing to improve the on-the-go experience so your knowledge stays usable beyond the desktop. Would love to hear what mobile workflows matter most to you!

@devagyasharma Figma grids vanish the moment you're in the browser, so I like that Grid Overlay Pro has Responsive Breakpoints + Quick Presets. Does it snap the overlay to a selected container, not just the viewport? That keeps reviews from turning into pixel debates.

@devagyasharma Congrats Devagya on the Launch!!

@artinbogdanov congratulations

@jeremy_ellis1 you have invite code?

@marti_serra_molina Thanks! That's the goal:)

@rohit_waghire It deploys a new server and sets up OpenClaw after the user buys the service. There is a Free plan, where you can run the script on your own server.

@victor_eth really good question, regarding crawling, staleness-based snapshots, not real-time. Each page has a refresh frequency (daily by default). We only re-crawl what's actually stale and skip re-indexing if content hash hasn't changed. Regarding rate limits, 100ms delays between requests, max 5 concurrent jobs per project, depth/URL caps and we identify ourselves with a custom User-Agent. Regarding robots.txt, yes, we respect it. We check for Sitemap directives first and prefer using the site's own sitemap over recursive crawling

Would love to hear more about your use case!

@selfishprimate Thanks Halil! The fun factor was intentional — I got tired of staring at terminal logs and JSON all day. Figured if I'm going to monitor AI agents, might as well enjoy the view. Let me know if you give it a spin!

@asadatnoodle Ha! I'll take that as a compliment. Though unlike Black Mirror, these AI agents actually work FOR you, not the other way around. And you can pull the plug anytime — it runs on your own hardware. Would love to hear what you think once you try it!

Good one! That was the first thing that came to my mind. 🤣

@benjamin_shafii Yes, we are trying to make an amazing experience both for non-technical users and devs. For devs, we are working on improving our SDKs and APIs, trying to make the data quality as high as possible and the performance really great, while providing good documentation. We are trying to listen more to talk to developers every day to improve the experience, and would love more feedback from you so we can make it really like an amazing integration with OpenWork.

https://docs.screenpi.pe/sdk-reference

@byalexai we do PII removal to remove risky elements and rigorously test the code

all data is local so it's as secure as other files on your computer

we have encryption at rest on the roadmap (eg. bitwarden / crypto wallet style locks) and support right now end-to-end encrypted device sync

also it's open source so you can audit the code or fix it

https://github.com/mediar-ai/screenpipe

@djkgamc We've released recently an Obsidian integration, but it's just Markdown files. You can configure it to run at a schedule (e.g., 15 minutes, 1 hour, 6 hours) and it would query your screen activity, microphone activity, keyboard mouse everything from the time zone time range you define and then it would give you this reflection. You can customize the system prompt and this under the hood is using a coding agent similar to Claude Code, so it can edit files, is read files and it's secure. It's like it's not just one shot prompt dump like request to AI and so it can properly extract like to-do lists and stuff like this.

@valeriia_kuna Actually it's aimed to be easy to use for non technical users. You can just run the app and ask AI questions through screenpipe AI chat or Claude integration

@janschutte it works great with just text LLM - we capture OCR, accessibility, so don't really need vision, but you can still process some frames to double check information

@daniele_packard All data stays on your computer by default, we use native Apple, Windows OCR and voice activity detection, segmentation and transcription

We also have feature to remove PII from OCR, DOM, transcriptions and screenshots like email, credit cards, etc. heavily tested

Uses 1-3 gb of RAM, 20 gb/m, 30% CPU