PH热榜 | 2026-06-27

Hi Product Hunt ! I'm Aymeric, co-builder of Folio 👋

Making slides is a huge time sink. And the current "AI for Slides" tools are lame: they make you wait for minutes and lost focus, and produce broken output.

We built Folio as the first real-time AI copilot for slides. It's over 6x faster than competitors, and makes high quality slides that you can iterate on. It supports native charts, all kinds of shapes, LaTeX equations.
It also works directly from your layouts and assets, to preserve your brand
Here are examples of produced by Folio in one shot:
- Air France 2025 financial statements
- 2025 State of AI by McKinsey

We've been testing over the past few months with power users form consulting or finance teams. The feedback has been incredible. Most are saving hours on slide-intensive days, one user told us it's like a ChatGPT moment for slides.

What makes Folio good? At the core of our tech is our slides engine: we have found a great way to represent and edit slides for an AI agent, combining both simplicity and expressivity, to burn 6x fewer tokens. We benchmarked it against leading slides copilots : it takes top rank, by a wide margin.

🎁 We made an exclusive promo code for Product Hunt: HUNT4FOLIO gives you one free month of Plus, on us!

Try the full version right now on Google Slides, or the demo on our homepage.

We're looking forward to your feedback to improve the product! 😃

Congrats on another great launch! One thing I'm trying to picture - when I say "make the chart bigger" and it stops fitting the template grid, do you guys push the neighboring shapes? Or hold them as constraints?

Huge congrats for shipping this @aymericroucher qq. on the native charts can Folio link directly to live data in Google Sheets or Excel, or does it generate static charts?

@aymeric_roucher I realize I might be spamming a bit, apologies. Quick additional question I had - what about web search? Asking as I’m using LLMs for research on the side, being able to prompt something like « check this website and update financial figures with the latest from last quarter »?

"we still get looksmaxxed on slides a little but we IQmog hard now." - @Claude by Anthropic

C'est vraiment assez impressionnant. À quel point on va pouvoir l'utiliser sur Google Slides, par exemple sur PowerPoint, je ne me rends pas trop compte. Mais je pense qu'en tant que consultant, ça peut vraiment changer ma vie. Merci beaucoup

Perfect timing on this launch. I'm literally building a fundraising deck for my startup's board this morning and I'm now thinking about switching over to try Folio mid-process.

The surgical in-place edit model is the thing I've always wanted. Every other AI slide tool I've tried feels like handing off your work to someone who comes back with something completely different. The "iterate without regressing" problem Mustafa mentioned is real and it's what made me give up on most of these tools.

it seems cool! is it better than gamma?

How does it compare vs Claude Cowork?

Great achievement Aymeric, can Folio work with an existing template ? I have been struggling with that lately

the iterate part is the thing that always breaks with AI slide tools. they generate something glossy on round 1 and then the second prompt regresses the whole layout.

how does folio handle rounds 2 through 5? if i say "make the chart bigger and move it left" does it preserve the rest, or does it regenerate everything and i have to redo the parts i liked?

that's been the deal breaker on every AI slide tool i've tried this year.

You mentioned optimizing your slide engine to bypass standard, bloated JSON/XML trees. When a user prompts for a complex structural change (e.g., "convert this 3-column layout into a 2x2 grid"), how does your simplified representation layer convey that structural shifting to the agent without exploding the token count? Is it using an abstract, markdown-like layout schema, or are you passing localized coordinate deltas?

the 6x speed claim is interesting — the bottleneck with most AI slide tools isn't generation quality, it's the iteration loop. you generate something decent, then spend 20 minutes manually fixing layout issues the AI created. does Folio handle the "fix what the AI broke" step too, or is it mainly about the initial generation speed?

The "works directly from your layouts and assets to preserve your brand" line is the part most slide-AI tools fake by dropping text onto a generic template, so the native charts and shapes claim is what actually makes this interesting. When it runs in PowerPoint, is it a local add-in acting on the open file, or does the deck get uploaded to your backend to process - and does it read my existing master/theme on first run so brand styling isn't a separate setup step? And when it iterates on a chart, is it editing the real native chart object so it stays editable, or swapping in an image?

The PowerPoint workflow is still one of the most painful places for business teams. The real value is not just generating slides, but getting from messy thinking to a presentation that has structure, hierarchy, and a clear decision path.

I’d be curious how Folio handles source material: does it preserve the user’s argument and evidence, or does it reshape the narrative? That distinction matters a lot for consultants, founders, and operators using AI for serious decks.

@aymericroucher hello, The in-place editing part is what makes this interesting for me. Most AI slide tools are good at generating a first draft, but real slide work is usually about small edits, layout fixes, charts, and keeping the brand intact. Working directly inside PowerPoint and Google Slides feels much closer to how teams actually build decks.

The "works from your layouts and assets to preserve brand" line is the real hook here. Most AI slide tools generate something that looks like a template fever dream. The moment you inject into an existing brand system rather than starting from scratch, the output actually survives a review from a real designer.

Curious how you handle edge cases where the brand guide conflicts with what the AI wants to do layout-wise. Does it defer to the existing deck structure, or override spacing to fit the content?

Amazing product ! Does it integrate well with the Office 365 stack ?

Looks great!

I wonder if the agent is able to read notes and comments on the slides? Ideally I’d like to use it « like a junior », where I do batches of reviews and and it corrects afterwards

Hey everyone, Charan from QApilot here.

Every QA team knows what they want to test. "Log in, add an item to the cart, check the order confirmation." Saying it takes ten seconds. Then the team spends six months turning that sentence into something a machine will run: step definitions, selectors, automation glue, flaky scripts that break on every UI change.

The intent was never the hard part. The execution layer was the tax.

CoWork takes your existing natural-language or BDD test cases, converts them into structured BDD/Gherkin context, builds an execution plan, and runs the test on a real mobile device.

When the app behaves differently, like a changed label, unexpected popup, or interrupted flow, CoWork replans and asks for human approval before moving ahead. When it needs input, like an OTP, it pauses instead of guessing. If it can’t proceed, it fails honestly instead of faking a pass.

Who it’s for: QA leaders, SDETs, mobile engineering teams, and product teams with existing test cases but low execution coverage before releases.

Common use cases: Regression testing, Release readiness, checkout/login flows, OTP-heavy journeys, app flows that break often, and teams trying to reduce manual execution without rewriting everything from scratch.

What makes CoWork different is the balance: AI execution where it can move fast, human control where judgment matters.

If you run mobile tests, I’d genuinely love your take. Try it here: https://qapilot.io/product/cowork

Thanks for being here for the launch. I’ll be in the thread all day reading every comment.

-- Charan Tej, Product Guy @QApilot's CoWork

Hi everyone! Aakash from QApilot here.

I'm one of the makers behind CoWork, and I'm really excited to finally share what we've been building!

CoWork helps AI and humans work together to make test automation more reliable instead of chasing full automation. We'd really appreciate your support on Product Hunt today, and I'd love to hear what you think. Thanks for checking us out!

Hey builders 👋

I am one of the makers behind CoWork!

We built CoWork - an AI agent that drives real mobile apps step by step, planning each action, reading the screen, and adapting when things go off-script.

We kept hitting the same wall: mobile automation breaks the moment a layout shifts. So we wanted something that actually sees and reasons about the screen instead of leaning on brittle scripts.

We'd love your feedback and support. Drop your thoughts/questions in the comments below and one of us will reply!

Is agent behaviour humanlike? Or it feels more like an executioner? Congrats on the launch!

@vidushee_geetam The idea makes sense technically, but I’m interested in the operational impact.

If a team already has a mature manual regression process, what changes after six months of using CoWork? Is the biggest improvement shorter release cycles, higher regression coverage, fewer flaky tests, or simply less engineering time spent maintaining automation?

I’d be interested to know which metric your existing customers see improve first because that’s ultimately what teams will justify the investment with.

@charan_tej_kammara The biggest challenge with mobile automation has never been writing the first version. It’s what happens after the app changes for the tenth or twentieth release.

You mention CoWork can replan execution when screens, labels or flows change, while still asking for human approval before changing intent. How much of that actually happens automatically in production? For teams using this every sprint, what percentage of failures end up needing manual intervention versus successfully recovering on their own?

That number would tell me more than any feature list because maintenance is usually where automation becomes expensive.

Congratulations on the launch, @charan_tej_kammara @jsurendranathreddy @vidushee_geetam !

Love the framing that the real bottleneck isn't writing test cases and agreed it's the execution layer.

Curious: when CoWork replans after encountering an unexpected UI change, how do you prevent it from "hallucinating" a valid path and accidentally masking a genuine regression instead of surfacing it?

Congrats on the launch guys!!

Nice idea! Is the generated automation framework-agnostic, or does it target a specific testing framework?

The selector breakage problem is real.... we've seen entire automation suites go down after a single navigation update.

Curious how CoWork handles dynamic test data mid-execution. if a flow requires an OTP or an auth password during the run, can that be injected in real time or does everything need to be pre-configured before the session starts?

the six-months-to-automate part is real. my team gave up automating our iOS app last year because every release broke the selectors. went back to manual regression and shipped slower for 6 months.

what i'd want to know about qapilot: when an iOS update changes the entire navigation hierarchy, does the AI re-plan or does it break the same way our selectors did? that's the test for whether this actually saves a QA team or just shifts the breakage to a different layer.

When scale testing mobile applications, how does CoWork handle deep state synchronization or state flakiness across parallel test instances? Is it actively maintaining a synchronized virtual state tree across the instances, or relying on aggressive DOM/view re-verification loops to prevent false negatives?

The replanning loop is what gets me here — most AI test tools fail silently when the UI changes, and you end up with flaky tests nobody trusts. Human-approved replanning before re-execution is the right call.

Curious how CoWork handles cases where the AI's planned steps diverge significantly from the original test intent — does the human reviewer see a diff of the original vs proposed plan, or just the new steps in isolation?

Congrats on the launch 🚀

The replanning-on-UI-change part is the real claim here - most mobile suites die exactly when a label moves or an unexpected popup shows up, so an agent that recovers is genuinely useful. When CoWork replans around a changed label, does it persist the adapted step back into the BDD/Gherkin definition so the next run is deterministic, or does it re-infer the path every run (which would make pass/fail non-reproducible across CI runs)? And does execution happen on a hosted device farm or on my own connected devices - that matters for builds behind auth or internal-only distribution.

Mobile QA is a strong place for automation because the bottleneck is rarely one test; it is the repeated device, environment, and regression coverage that slows teams down.

The “same QE team, 3x automation” positioning is interesting. I’d be curious how QApilot handles flaky tests and app-state changes, because reliability is usually what determines whether QA teams trust automation or keep falling back to manual checks.

@charan_tej_kammara The “fails honestly instead of faking a pass” part is the strongest detail for me. In QA, a test that silently adapts in the wrong direction can be more dangerous than a broken test. Human-approved replanning around the original test intent feels like the right balance between speed and trust.

looks cool! can it work on my local device?

How would this compare to doing UI automation testing with a tool like Selenium or CypressJS

Congrats on the launch! 🚀

I like the human-approved replanning approach. Mobile automation often breaks on small UI changes, popups, so having AI adapt while still asking for approval feels like the right balance.

Curious how CoWork handles flaky test behavior across different real devices and OS versions.

Wow. Nice one. Do it have the same operational function for web? How variety of the test cases this cover?

The "human-approved replanning" step is the detail that makes this production-viable. Most AI automation tools pitch full autonomy, but that is exactly where QA teams lose confidence in the output. Having the AI propose the plan change and a human sign off before execution means you get the speed gains without the "the bot went rogue on staging" stories.

How does the replanning loop work when a test fails mid-run? Does it pause for approval every time, or only when the deviation is above some threshold?

The human-approved replanning is the right call for mobile. One question: when a test needs to switch between apps mid-flow (e.g., grab an OTP from email, then switch back), does CoWork handle that as one session, or does cross-app context reset?

Last November, my friends and I started thinking about how we could make it easier for musicians to create music anywhere, anytime.

At the time, we watched the making of video for Stromae’s “Alors on danse.” It was funny, but also inspiring. It made us wonder: what if people could create music using only their voice?

A lot of people have great musical ideas, but they cannot always play an instrument. Traditional DAWs can also feel too complex, especially when all you want to do is capture an idea quickly. But almost everyone can hum, sing, or whistle a melody.

That was the moment everything clicked.

We decided to build Nada, which means “tune” in Indonesian. At first, we imagined it as a better voice notes app for musicians, something simple for capturing musical ideas. But as we kept building, it slowly became a mini DAW: voice-to-MIDI, instruments, arrangement, beats, and music production features, all designed to stay simple and not overwhelming.

Today, we’re excited and proud to finally share Nada on Product Hunt.

We built Nada for musicians, creators, and anyone who has ever had a melody in their head but struggled to turn it into music.

We’d love to hear your feedback, ideas, and thoughts. And if you’re a musician, creator, investor, partner, or someone who believes music creation should be more accessible, we’d love to connect.

For more inquiries, you can contact us via email: tunelabid@gmail.com

Thank you for checking out Nada!

For the best experience with Nada, we highly recommend using wired earphones with a built-in microphone, such as Apple USB-C EarPods or any similar wired headset.

Wired earphones help produce better-quality results because they reduce latency and prevent speaker audio from interfering with pitch detection. Wireless earphones may introduce noticeable latency, while using Nada through speakers can reduce pitch detection accuracy.

For the most accurate voice-to-MIDI result, wired earphones are highly recommended.

If you haven't tried the Chord Pads yet, please give it a spin! It's one of our standout features, designed to help you generate chord progressions and kickstart your music ideas. Cheers!

It’s amazing to finally see Nada live on Product Hunt! 🎉 been watching this idea grow from a simple thought into a tool that helps people turn melodies into music has been such a journey. Really proud of what Tunelab team built. Hope Nada helps more creators bring their musical ideas to life.

The "no AI generation" framing is actually the right call here - the value is capturing your musical idea, not generating a new one. Musicians who can hear melodies but can't play piano have been stuck with voice memos + manual transcription forever. Cutting that to hum-to-MIDI directly removes 3-4 steps from a genuinely painful workflow.

Curious how it handles rhythm accuracy - is the MIDI timing quantized automatically or does it preserve your exact phrasing? That's usually where voice-to-MIDI tools live or die for people who care about their groove.

Hi everyone! I’m Tika, part of the design team behind Nada. 👋

From the very beginning, we knew we wanted to create something that lived at the intersection of creativity and productivity. Since we all share a love for music, it naturally became the heart of what we wanted to build together.

Designing this app has been both incredibly fun and challenging, and I feel privileged to have been part of it since day one. I’ve watched Nada grow from messy whiteboard sketches into an intricate app that continues to evolve with innovative features we never even imagined at the start. It has become a product we’re all genuinely passionate about.

What makes this journey especially meaningful isn’t just the product itself, but the team behind it. I’m deeply grateful to be part of this amazing group of people and excited for what lies ahead for Nada. We believe there’s still so much to explore in refining the creative process for musicians, and perhaps even embracing AI to unlock new ways of creating, capturing, and shaping musical ideas.

This launch feels less like the finish line and more like the first note of a much longer song. 🎵🚀

Nada Coder here. I really enjoyed building this product and felt excited throughout the entire development process. From the very beginning, during the ideation phase, our goal was to create a product that is not only enjoyable for us to build as developers, but also enjoyable and valuable for users to use in their daily experience. We continuously focused on understanding user needs, collecting feedback, and improving the product based on the experiences and insights we received along the way.

Throughout this journey, I learned a lot because this product was something completely new to me and pushed me to explore many areas that I had never worked on before. During the development process, I encountered various technical challenges, unexpected issues, and many moments where I needed to learn quickly in order to solve problems effectively. Those experiences helped me improve not only my technical skills, but also the way I think about problem solving, product development, and collaboration within a team.

What made this experience especially valuable for me was seeing how an idea can slowly evolve into a real product through continuous iteration and learning. Building this product taught me that creating good software is not only about writing code, but also about understanding users, adapting to feedback, making improvements consistently, and being willing to learn from mistakes. Overall, this journey gave me a lot of new knowledge and became an experience that helped me grow significantly as a developer.

Hi product hunt! 👋

I’m one of the designers behind Nada. It’s been really fun watching this app grow from early designs into something people can actually use.

I’ve mostly been involved in making the product and visuals feel easier to understand and more enjoyable to use. Seeing people turn a simple hum into something they can build on has been one of the coolest parts of working on this project.

Excited to finally share it here, and I’d genuinely love to hear what you think. Thanks for checking out Nada! 🫶

For the voice-to-midi translation engine, how are you separating expressive articulation (like pitch bends, vibrato, or dynamics) from the raw musical notes? Are you mapping those vocal nuances directly to standard MIDI CC data in real time, or handling them post-processing through a secondary latent-space conditioning layer?

Love that this is "capture your own idea" and not "AI makes a song for you." Great work!

@bregas This would honestly save so many random voice notes on my phone. The idea of humming something quickly and turning it into MIDI before the melody disappears feels super useful.

the dock for classic mac OS 9 is the unlocked badge of every developer who's quietly been wanting this since they were 14. shut up and take my money.

real question though: does the CRT shader work with external monitors? half my screen real estate is at 27" and i don't trust most shader presets to scale clean across resolutions.

Hey, I saw this relaunch and immediately saw something good was coming (and I was right)!

Anyways, whens the windows version coming...

I love these Retro series and now incorporated the Simpsons! :D You are a genius!

Congrats on the launch! 🚀

The attention to detail is impressive, especially recreating the Windows 98/XP taskbar behavior instead of just changing the visuals.

I'm curious: since the shaders run across the whole desktop, how much GPU overhead do they add during everyday work or video calls? It would be interesting to know how they perform on older Intel Macs versus Apple Silicon.

This launch got me :)

From a systems-level perspective, how are you intercepting the window server/compositor to inject the vintage display rendering and retro behaviors without triggering major layout lag or shattering native macOS window-snapping frameworks?

This is genuinely delightful. The CRT shader on a single window is a nice touch — you can keep your workflow normal and just make the one app you're vibing with look like 1987.

Curious whether the shaders work with external displays or if it's limited to the built-in screen. And is there any performance hit on M-series chips running the full-screen presets?

Solid free tier strategy too — let people fall in love with it before asking for money.

@maik_klotz This is the kind of app people open “just to try once” and then accidentally spend an hour changing their whole setup. The virtual webcam part is a fun touch too — showing up to a call with a CRT look is weirdly tempting.

I urgently need this.

Hi everyone! I'm Kevin Brown, one of the makers of Cloud World Model.

Cloud World Model lets you model AWS, GCP, Azure, OCI, and DigitalOcean architectures and instantly see how they behave CPU, error rates, throughput, autoscaling, failure recovery, and cost without provisioning a single real resource.

A few things we're proud of:

• A capacity-aware engine that models real per-provider performance profiles

• Chaos engineering: inject zone outages, DB crashes, and network partitions, then get a resilience score

• A multi-cloud explorer that compares provider combos on cost, latency, and vendor lock-in

• A full RL training API so AI agents can learn cloud optimization in a safe, cost-free environment

• Beginner mode with plain-English AI explanations and an interactive tutorial

Whether you're learning cloud skills or training agents to optimize infrastructure, I'd like to hear any of the following in the comments?

1. How do you typically test cloud architecture changes before putting them in production or any environment?

2. Do you think a mechanism to be able simulate a cloud architecture change would be useful?

3. Any experiences with cloud cost comparisons?

When simulating highly stateful infrastructure setups (like managed DB clusters with strict VPC networking rules or IAM permission chains), how deeply does your local mocking layer mirror the cloud providers' internal API state validation? Does it execute structural validation against a custom internal schema engine, or parse translated Terraform/CloudFormation configurations directly?

Really cool - had a multi-cloud setup simulating in a couple minutes. One thing though: connecting resources took a few clicks each time (here's what I mean: Cloud World Model | createademo, ~0:10). Is there a faster way to wire them up - a drag-from-handle or keyboard shortcut? Would speed up building a setup a lot. Congrats on the launch!

Useful angle for teams that want to teach cloud tradeoffs without handing out real cloud accounts. I’d be interested in how close the cost/perf model stays to provider changes over time, since drift is usually where these simulators get hard to trust.

This is the one I keep coming back to, cost is the question that never really leaves the room, and almost always the hardest thing to pin down before you commit.

My real question is fidelity. The headline compute numbers are easy, every calculator gets those right. The bills that actually blow up are the hidden line items: egress and cross-AZ traffic, managed-service markups, spot vs committed pricing. Does the engine reach down to those, or just the sticker compute price? And the case that would really earn its keep: migration, where the egress to leave a provider ambushes everyone and never shows up in a "provider A vs B monthly" comparison until the invoice lands. Can the explorer model the transition cost, not just the steady-state side-by-side?

Genuinely love the concept, the chaos-engineering resilience score is a great touch too. Congrats on the launch! :)

@mathsociety I like this because cloud mistakes usually get expensive after you already deployed them. Being able to play with failure scenarios, cost, and scaling before touching real infra feels especially useful for learners and small teams that don’t have a proper staging setup.

The RL training API is the sharp end of this — and the part I'd push on. An agent is only as honest as its reward. Cost, latency, and resilience are in tension: minimize cost hard enough and the agent learns to ship something cheap and brittle that looks great right up until the zone outage you didn't simulate.

So, two things I'd want before trusting an agent's infra recommendation enough to act on it:

Is the reward multi-objective and user-weightable (I decide cost vs resilience vs latency), and does a run surface the tradeoff the agent chose — "cut 30% cost but dropped your resilience score from 8 to 5" — instead of just handing back one "optimal" config? The tradeoff being visible and mine to set is the whole ballgame.

The chaos-injection + resilience score framing is a great call too. Congrats on shipping, Kevin

Congrats on the launch! 🚀

Simulating cloud architecture before provisioning real resources is a very useful idea, especially for cost-heavy experiments and failure testing.

I'm curious: how close are the cost and performance predictions to real-world cloud bills after deployment? Do you provide any confidence score or comparison against actual usage data over time?

The RL training API is the part that grabs me - an agent is only as good as the sim it learns in. The capacity-aware engine modeling "real per-provider performance profiles" is where that lives or dies: are those profiles grounded in published benchmarks and vendor specs, or in measured telemetry, and how often do you refresh them? If the sim cost/latency drifts from the actual providers, an agent will happily optimize for the model instead of the cloud, so how do you validate fidelity against a real deployment?

Strong launch. The RL training API is the interesting edge. If an agent learns an infra optimization in simulation, I’d want the handoff receipt before deploy: resources changed, env/secret assumptions, failure case tested, and rollback path.

Do you expect agents to export a plan into Terraform/Pulumi, or stay inside the simulator?

the cost simulation is the part i need most. i blew $400 on an RDS instance i spun up for "testing" and forgot about for 11 days. nobody warned me.

how granular does the cost projection go? if i model a 3-tier app does it tell me i'm about to pay for an over-provisioned NAT gateway, or just give me a total bill estimate?

the value of cost tools breaks for me at the line item level. that's where i actually make decisions.

The chaos engineering part caught my eye, injecting a DB crash and getting a resilience score back seems really useful for catching weak spots before prod. Curious how close the cost estimates land to a real AWS bill in practice. Congrats on shipping!

great launch! this looks really interesting. as an academic writer, I’m always bouncing between docs and apps while writing, which is okay for single manuscripts but tough for books or monographs…I’ve got an upcoming project and I might give this a try…how does it compare to an app like scrivener (other than cost)?

Oh, I felt this one personally. Twelve docs, character notes in three places, feedback living somewhere in the void.

Love that you’re solving the boring-but-deadly part of writing: keeping the book alive while everything around it tries to turn into chaos.

How Epilogue handles messy second drafts? when the story is already there, but half of it needs to merge or die.

This market (serious author writing apps) is genuinely crowded - Scrivener, Ulysses, iA Writer, even Obsidian with the right plugins. All of them make the same "no distractions, keeps you in flow" promise.

The alpha reader sharing angle is the one thing here that's actually underserved. Getting a draft in front of 20 beta readers without turning it into a Google Docs mess is still painful, and nobody's solved it cleanly yet.

What's the concrete differentiation from Ulysses specifically? That's probably your closest direct competitor for the "offline-first, focused, serious author" positioning - and the answer to that question is what would make me switch.

Wow, keeping alpha reader feedback private really helps one loud note not anchoring the whole group! Most writing tools get this backwards by defaulting to public comments... Love the product, you seem to be on the right path

Isn't this something similar to Elipus?

the sneaky part of mixed-fps review: a 24fps clip and its 60fps interp share an instant only every 1/12s, so stepping both one frame drifts them apart. nearest-pts snap on the slower clip keeps it honest.

Love the idea of keeping reviewing separate from editing. There are plenty of times I just want to compare outputs, not open a full NLE project.

One workflow I'd find really useful is comparing multiple AI-generated versions side by side (upscaled, restored, interpolated, etc.) and being able to bookmark specific frames where one model performs better than another. That would make model evaluation much faster.

Frame-accurate multi-cam sync via audio analysis is the feature that makes this actually useful for real production work. Syncing multi-cam footage manually is one of those jobs that silently steals 30 minutes from every edit session - doing it automatically with a single click is the right call.

The 12-panel limit covers any realistic multi-cam setup. One edge case I'd be curious about: does it handle mixed frame rates side by side (say, 24fps B-roll next to 60fps slow-mo)? That's where most video comparison tools fall apart. Congrats on the launch.

Hey! 👋 Girish Kotte here, Founder of Wysera.

The honest origin story: we were running three separate tools - one for content, one for CRM, one for ops automation and spending more time keeping them in sync than actually growing. Every handoff between tools was a gap where context got lost, deals slipped through, and the brand voice drifted.

That frustration became Wysera. The core insight was simple: if the same AI that writes your blog post also knows your pipeline, your open deals, and your customer history, it stops being a content tool and starts being a genuine business co-pilot.

Here's what that looks like in practice:

→ PostWyse drafts SEO content - but Wyse also flags which topics map to your stalled deals

→ OpsWyse tracks your CRM - but the same agent can write the re-engagement email right there

→ Every approve/dismiss you click teaches Wyse your brand voice and decision logic over time

We're early and we know it. But we've already seen teams cut their tool stack from 5-6 point solutions down to one feedback loop.

If you're a founder, marketer, or ops lead who's tired of duct-taping AI tools together - I'd genuinely love to hear what your stack looks like today. What's the one workflow you wish just worked without you babysitting it?

Happy to go deep on anything - AMA! 🙏

@gkotte I noticed one of your promises is “Approval is the default,” and I think that’s an important design decision.

A lot of AI products push for full autonomy, but most businesses aren’t comfortable letting an AI send emails, update CRM records, or change customer communication without oversight.

As customers build confidence, do you see most teams eventually moving toward autonomous workflows, or do they generally keep approvals in place for customer-facing actions even after months of usage?

@gkotte I think the shared context is the strongest part of the pitch, not replacing HubSpot or Mailchimp.

Most AI tools today are really good inside their own lane, but they lose context the moment you move from content creation to sales or customer success. If the same AI actually understands what was published, which leads engaged, which deals stalled, and what follow-up makes sense next, that’s a different category of product.

My question is where you’ve found the practical boundary. At what point does one shared context become too broad, and when do you decide a workflow is better handled by a dedicated specialist instead of the central agent?

Congrats on launching! 🚀 Replaces HubSpot, Salesforce, and Mailchimp? Big promises. The design looks super clean, but my instant reflex when seeing a product claiming to automate everything from top-of-funnel SEO to pipeline closing is heavy skepticism. In reality, 'all-in-one' AI tools usually write generic, bot-like copy, trash your email deliverability, and turn your CRM database into a chaotic mess that a human has to spend hours cleaning up. How do you prove this isn't just a giant, brittle wrapper over a bunch of basic prompts that I’ll end up babysitting anyway?

The shared context angle is the real pitch here and it's more defensible than "replace five tools." The idea that the agent drafting your SEO content also knows which deals went quiet is a legitimate insight - that's not something a Zapier stack gives you. My skepticism is about maintenance overhead: when content, CRM, and ops each evolve their workflows, a unified agent has to absorb all that change simultaneously. Point solutions let each team ship independently. What does the experience look like when one department wants to change how Wyse works for them without breaking the context the other teams depend on?

@gkotte How are customers actually adopting Wysera today? Are companies replacing their stack all at once, or are they starting with PostWyse or OpsWyse and gradually expanding once they trust the shared context?

The "every department" angle is ambitious, most AI agents solve one workflow well and call it done. Curious how Wysera handles context switching between departments; does each agent operate independently or do they share a unified business memory?

@gkotte , Nice launch! The shared context across content and CRM is the part that stands out, that's the gap a Zapier stack never really closes. One thing I'm curious about: how does Wysera handle brand voice shift across departments? When the same agent writes a blog post and a re-engagement email, do teams find the voice stays consistent, or do they end up tweaking it per surface? Following along.

I'm one of the builders behind OpenBot. We're incredibly excited to launch today after months of work.

OpenBot helps you create specialized AI agents with natural language and coordinate them on tasks like teammates.

Congrats on the launch Guys!!

Btw, Does OpenBot support long-running agents and persistent memory across sessions?

Congrats on the launch @ddaras @giorgi_daraselia !

What is the first specific use case you want new users to try and immediately understand the value of OpenBot? What makes OpenBot different than Manus?

love this guys congrats @giorgi_daraselia 👏web app stack example is perfect. qq how long did it take you to configure OpenBot to handle that specific Firecrawl -> Stitch -> Claude pipeline from scratch?

Which agent is mostly used?

This looks very interesting, will give it a try later next week. Congrats on the launch, looks like a top product

We built Velocity for the people who don't settle for Average AI results.

It's a Chrome extension, that's already saving users 1000s of hours & reducing token wastage for longer, more meaningful sessions with AI. All this while also improving output quality, instantly!

Velocity uses an intent-driven architecture that understands your goal and turns messy text & rough thoughts into optimised context-aware prompts in one click (within the text box).
Resulting in consistently better outputs across ChatGPT, Claude, Lovable, Runway and 40+ AI tools.

Our early user's and enterprise clients say, "it's simply unfair". Try the Web-App or explore the GOLDMINE of Ready-to-use Prompts in the Prompt Library.

We approach a future where agents do our work, so the only differentiator will be the human in the loop and their capabilities. So we built this. In a nutshell, it's Grammarly for Prompting but, on steroids. It doesn't just enhance prompts, it enhances your thinking with features like contextual refine & suggestions. With it's unified memory, it remembers & learns from your usage across any AI tool, so you don't have to repeat yourself!

The truth is Better prompt = Better Outputs. Velocity makes sure they're great consistently. It's domain and model agnostic so anyone who uses AI seriously, benefits from this tool. We'd love to hear where you see the most gains!
It's quite straight forward, You Think it Velocity makes sure the AI understands it ;)

PS: We have a referral system allowing you to GET 1 MONTH VELOCITY FREE!
In case you run into any issues, Mail us at Velocity@toteminteractive.in or dm me on Linkedin and we'll get you sorted!

Huge congrats on shipping this @aakashpuri , qq Is there a way to set a default system prompt or brand voice that automatically applies to every chat session?

actually useful and the memory feature is a nice touch.

When does the Desktop app come out!??

Hey Product Hunters thank you for checking out A.S.S.H.O.L.E.™ 😃

It turns out the dumbest ideas are sometimes the most useful.

Yes, the acronym was absolutely intentional.

I created this as a parody of the AI industry, where every week seems to bring another "revolutionary" assistant that's supposedly going to organize your life, optimize your workflow, and solve all your problems.

That said, I'd be lying if I said I wouldn't actually use this.

An AI that remembers everything, challenges bad decisions, reminds me what I forgot, and occasionally says, "That's a terrible idea." Sign me up.

I'd genuinely love your feedback:

• What's your favorite fake feature?

• Would you actually use an AI that's brutally honest?

• And what should A.S.S.H.O.L.E.™ do next?

Thanks for stopping by, and if this made you laugh, then the launch was a success. 🍻

FYI - my entire video was made using Visla's Director Mode feature! Pretty cool!

Respect the self-awareness here - the parody framing is genuinely funny and the critique of "revolutionary assistant" launches is fair. My honest concern is that the joke and the product are doing different jobs. The name gets you traffic on day one but makes it almost impossible to recommend to a colleague without a long explanation. "You should try this tool, it's called A.S.S.H.O.L.E." is a tough sell in most work contexts. The actual use case - an AI that challenges bad decisions and remembers what you said - is real and worth building. Just wonder if the branding becomes a ceiling rather than a launchpad.

Congrats on the launch, Hemant! I like that you're focusing on behavior patterns instead of just streaks.

One feature I'd love to see is a "habit recovery" view. Everyone misses a day occasionally, but what's often more important is how quickly they get back on track. Seeing recovery time alongside consistency could give users a much healthier picture of long-term progress than streaks alone.

The post-launch use case you described is the more compelling angle honestly - the "agent wrote code but doesn't know what changed between releases" problem hits every team running AI-assisted CI. Pre-launch gap detection is useful once, but version drift and postmortem tracking is a recurring pain. Does VibeRaven watch git diffs automatically, or do you trigger the scan manually after each deploy?

If you have any questions or want to just share things, feel free to send me a message on LinkedIn

I like that you're shifting the goal from simply blocking apps to creating a healthier habit before opening them.

One feature I'd personally enjoy is adaptive sessions. If I'm about to unlock an app for the fifth time today, the app could gradually increase the yoga session from 30 seconds to a couple of minutes. That keeps the friction proportional without making it frustrating from day one.

Congrats on the launch! The page-cited chat is a nice addition since it makes it much easier to verify where an answer came from.

One workflow I'd love to see is exporting structured data directly from documents for example, turning invoices, tables, or forms into clean JSON or CSV after the OCR step. That would make it useful not just for reading documents, but also for automating data extraction pipelines.

Congrats on your launch ! Is the OCR proprietary?

The pain point is so real - arriving at a library only to find every outlet taken or the wifi password has changed is genuinely frustrating. The data layer is actually the hard part here: how are you sourcing and keeping the location info current? User submissions, scraping, or manual curation? That'll determine whether this stays useful 6 months from now or drifts stale.

The core problem you're solving is real - agent workflows have a broken approval loop. You're expected to sit and watch for the 'can I proceed?' prompt, which defeats the whole point of delegating to an agent. A physical interrupt button that works while you're away from the screen is a legitimate answer.

My skepticism is on the desk pet framing. "Throws confetti for your wins" and the cheerleading layer feel like things that get turned off within the first week. The control mechanism is the durable value here - the personality risks being the thing that keeps developers from putting it on their desk at all. Is there a minimal mode for people who want the button without the buddy?

Hey Hunters 👋

It’s great to be back! A little over a year ago, We launched the first version of Themefy here. Our goal was simple: to take the mesmerizing world of creative coding and generative art and put it right into your browser's new tab. The response was incredible, and seeing so many users turn their browsers into interactive canvases was truly inspiring.

We also received some fantastic feedbacks and we listened, went back to the drawing board, and rebuilt the extension from the ground up. Today, we're so excited to introduce Themefy 2.0! 🚀

We’ve transformed Themefy from a simple background extension into a complete, premium new tab experience that perfectly balances aesthetics and productivity.

✨ What’s New in this massive update?
1. Immersive 3D & 2D Animations: We've leveled up the interactive backgrounds with stunning new 2D and 3D scenes and generative artworks.
2. Premium Wallpapers: Enjoy breathtaking photography.
3. Smart Productivity Widgets: We’ve added a beautiful glassmorphic Top Sites sidebar for quick access to your favorite pages, alongside an advanced multi-engine search bar.

We would love to hear what you think of the new update! What kind of themes or widgets would you like to see next? We’ll be hanging out in the comments all day to answer your questions and chat.

Enjoy! 🎨
— Bhuwan

New tab real estate is so wasted by default - just blank space you glance at for half a second. The 3D animation angle is genuinely different from the static wallpaper approach every other new tab extension takes. Does the theme library stay fresh automatically, or do you have to manually update? And curious whether settings sync across devices via Chrome sync or it's purely local.