PH热榜 | 2026-04-27

Hey everyone, Vihaar here, one of the founders of orange slice

really appreciate you checking us out today

we didn’t start this company because we wanted to build “another sales tool”

we started it because we kept running into the same problem over and over again

every go-to-market motion we tried (whether it was outbound, partnerships, hiring, events) always broke at the same place

finding the right people at the right time

not a big list
not more data

just… actual in-market companies

and the frustrating part was the workflow

copy pasting people into spreadsheets
running the same enrichments over and over
stitching together 10+ tools
hoping something sticks

it just felt… dumb

so we started building what we actually wanted:

a system where you can just describe who you’re looking for

and it finds them for you
researches them
enriches them
filters them
updates continuously

all in one place

over the last few months we’ve:

• rebuilt the product multiple times

• shipped a self-serve version after doing a lot of agency work

• broken things (a lot… memory in chat almost killed us last week lol)

• and most importantly, started seeing people use this in ways we didn’t even expect
  

some are using it for outbound
some for market research
some for partnerships
some for completely random workflows

that’s been the coolest part

we’re still super early

but if you’ve ever felt like go-to-market is way more manual than it should be
or that timing matters way more than effort

i think you’ll get what we’re trying to do here

would genuinely love feedback - good or bad

and if you want, comment your companies website

and i’ll reply with a custom GTM play
that will help you get more customers!!!

It is a good sighn that users are applying it beyond outbound. That usually means the system has some flexibility built into it.

Anyone NOT seeing this? First step of onboarding I'm told they cannot do this right now - on launch day...?

Congrats on the launch @vihaar_nandigala I believe the timing angle is underrated in GTM - most tools optimize for volume, not relevance.

How are you defining “in-market”? Is it behavioral signals, firmographic triggers, or something else?

Congrats on the launch @vihaar_nandigala !

When you say "automate any sales task," what's the most repetitive sales workflow your beta customers have fully offloaded to the agent?

And on the flipside, what's the sales task that still needs a human in the loop today?

Curious how this handles the research layer when you're running outbound across multiple e-com brands at once.

Do you pull brand-level signals recent ad spend patterns, new growth/CX hires, post-funding activity or is it mostly contact-level enrichment? Because when you're managing 10-15 stores the prioritization layer matters as much as the messaging itself.

Hi, a few questions -
1. How does the pricing work? If it's credits-based/usage-based, then how do you define that exactly?
2. How accurate is the data + how recent? What sort of data providers do you use? Especially intent signals are notoriously inaccurate for "things like people with 4 fingers as you say in YC demo" haha
3. How does it handle vague/intent based queries instead of traditional filter style queries? If I say something non-filter style like "find me users who are on the lookout for Instagram marketing automation tools"

Has been an awesome tool for building out lead lists and easily plugs into the rest of my tech stack. great founding team too

Been using orangeslice for a couple weeks and have really enjoyed it! As a gtm leader for a very small startup its a fast way to find leads and reach out to people in our ICP. @vihaar curious about seeding for linkedin/reddit I was messing around with those templates. I was hoping to get connected with someone from your team to help me maximize my capabilties of your product. Congrats again think what you are building is awesome!!

I wish they hadn't revealed this. This has been my secret weapon in finding the best targets for my company. I've built out some workflows that as SUPER niche and i swear OrangeSlice hasnt missed yet. 🥶

Love this direction. Pulling in real-time signals from across the web instead of static lists makes a lot of sense. What sources have been the most useful so far?

Congrats on the launch! Plain English ICP descriptions to a qualified lead list is a really clean workflow, no more Boolean wrangling. Curious how the system handles edge cases when the criteria gets really fuzzy?

Interesting take @vihaar

Feels like the hardest part here is actually figuring out who’s really in-market, not just matching filters

Curious how you’re handling that?

love the table interface. what was the decision around that vs chat-based?

This is brilliant.

Currently running GTM for two clients, and despite a lot of efforts, it still feels kinda janky.

Rooting for you guys!

Congratulations on the launch. Are you listening to other sources than reddit?

Hey everyone 👋

We built Jet AI Agents because most teams still:
- jump between tools
- manually run workflows
- rely on engineering for simple internal tools

So we asked:
👉 what if business teams could build their own AI agents — on top of real data — and actually automate work?

With Jet, you can:

- build ai agents without code

- enrich AI with your data & 200+ integrations

- let them answer questions and take action in Slack, Telegram, WhatsApp and more

- instantly generate visual reports

- self-host in your own environment

- use open-source AI models

- bring your knowledge into AI — from files, drives, websites, and multiple formats like DOCX, PDF, JSON, MP3.

We’ve also created a few templates to help you get started (the real magic is in customizing them ✨):
📊 Data Analysis Agent 👉 https://www.jetadmin.io/agent-templates/bigquery-data-analyst

🗓️ Meeting Preparation Agent 👉 https://www.jetadmin.io/agent-templates/meeting-prep-agent

🎧 Support Agent 👉 https://www.jetadmin.io/agent-templates/customer-support-agent

📝 Meeting Analysts Agent 👉 https://www.jetadmin.io/agent-templates/meeting-notes-agent

Would love your feedback:
👉 What’s the first workflow you’d automate?

The 'relying on engineering for simple internal tools' line is the story of my life. My backlog is 6 months deep with 'just one more internal dashboard' requests. If Jet lets my business ops team build their own BigQuery analyst without touching a line of code, you’ve just saved me 20 hours a week. Does it handle write-back permissions safely?
@anton_svetlov

interesting take on bundling workflows + agents under one platform. one thing i'd be curious about — when you have agents calling into the same data layer simultaneously, how do you handle scope conflicts? in my own setup (claude code on a nuxt 3 + go-zero stack) i ended up writing per-folder AGENTS.md files just so concurrent sub-agents wouldn't step on each other.

does jet handle that internally or is it more about the orchestration layer? curious about the design choice.

"AI agents your team will trust - because they built them themselves" hits different. we've had mixed results with off-the-shelf AI tools in healthcare workflows, but letting domain experts build their own makes sense. what's the learning curve like for non-technical users?

the Slack integration is smart - we've been looking for something that lets our team build agents without pulling devs away from core product work. curious how the 200+ tool connections handle auth and permissions? does each team member need to connect their own accounts or can you set up shared service accounts?

The knowledge ingestion piece is interesting. When you bring in docs from drives and websites, how does the system handle freshness? Is it a one-time import or does it stay in sync with the source? That feels like the part that makes or breaks trust in agent answers over time.

As a founder, I’m tired of asking my engineers to build simple internal automations. If I can connect our BigQuery data to a Slack agent without a ticket, it changes our speed entirely. Grabbing the PH exclusive now to test this out. @f1nal @Jet Admin

congrats on the launch. the "answer questions AND take action" piece is the part i'd want to grill when an agent has write access through 200+ integrations, how does jet handle action scoping ? does the business user who builds the agent have to whitelist specific actions per integration, or is it all-or-nothing per connection ? a poorly scoped agent firing into the wrong slack channel or messaging the wrong whatsapp contact is the failure mode that scares me as a builder.

This feels like a solid alternative to stitching together multiple tools. Especially for things like admin panels and client portals. How do you see this fitting into more engineering-heavy stacks?

genuinely curious where the 200 integrations thing breaks down in practice. Every tool says that and then you find out your specific tool is the one that doesn't work. what are the gaps rn?

Love this. Building agents that actually live inside Slack and WhatsApp instead of yet another dashboard tab is the right call. Curious which integrations teams reach for first when getting started?

Hey Product Hunt. I'm Steve, co-founder of Logic.

When you build an AI agent, the call to the LLM API is the easy part. The hard parts are evals, RAG, observability, prompt refinement, model selection, fallback, cost and latency tuning, system integrations, and giving the agent tools to do useful work in the rest of the world.

Logic gives you an out-of-the-box answer for all of that, while also improving how reliably your agents follow instructions.

With Logic, you write a simple spec that explains what the agent should do. We give you back a managed agent that can be called via MCP, REST, a web UI, or a dedicated email address. We generate well-typed schemas and synthetic tests, handle versioning, observability, and RAG, and give your agents a "batteries included" tool suite:

• Real-World Capabilities: All Logic agents can read 130+ document formats, fill out PDF forms, semantically search your knowledge library, send and receive email, do research, generate and annotate images, and call HTTP APIs.
  

• Smart Model Routing: Route across OpenAI, Anthropic, Google, and hardware-accelerated open-source models, with fallback and cost/latency tuning, so you can improve reliability without being locked into one provider.
  

• Deep Integrations: Easily connect to external tools like Linear, Notion, and any MCP endpoint.

We make your agents smarter.

When Logic's agent harness was measured against Allen AI's IFBench, one of the hardest public tests for precise instruction following, Logic scored 83.3% – higher than any model on the Artificial Analysis leaderboard. This is a six-point gain for the agent harness above the same base model (Gemini 3.1 Pro) when called directly.

So far, 250+ organizations have automated over 4M agentic tasks with Logic. Common use cases include things like content moderation, document parsing, data extraction, medical coding, and user onboarding.

Logic is SOC 2 Type II and HIPAA certified, there's a free tier, and paid plans that scale with usage.

Jess, my co-founder and CTO, and I will be in the comments. We're excited to see what you build with it, and we'd love to hear what else you wish it could do.

Thanks for taking a look.

routing across multiple providers is a strong feature. vender lock-in has been a real concern lately, so this helps reduce that risk.

The spec-driven approach is the right abstraction here. When I was CTO scaling from 15 to 120 engineers, the biggest pain with internal AI tooling wasn't the LLM call itself - it was everything around it: eval harnesses that nobody maintained, prompt versions scattered across repos, and zero observability into why an agent started failing on Tuesday. The fact that Logic handles model routing, versioning, and evals out of the box means teams can skip the 3-month infrastructure detour and actually ship. Curious how you handle spec evolution - when a team realizes their agent needs a fundamentally different approach mid-production, how smooth is the transition between spec versions?

The SOC2 HIPAA angle is important if this is truly targeting enterprise adoption.

Really like this direction. Turning plain English specs into production-ready agents is a big unlock. How are teams typically structuring their specs to keep outputs consistent?

Spec-driven agents with versioned rollback is rare. How much of the IFBench gain comes from the harness vs the synthetic test generation step?

model routing caught my attention. are you handling fallbacks automatically when one model is down, or is it more about cost/performance optimization? seems like a huge operational headache to manage manually across different providers.

the structured spec approach is interesting - how granular can you get with the agent behavior definitions? we've been building healthcare agents and the biggest pain is always the gap between "here's what it should do" and actually getting consistent behavior in production.

This is a big unlock for teams shipping agents. Writing a spec instead of stitching together prompts, retries, and eval harnesses sounds like a huge time saver. Any plans for letting teams share or remix specs across orgs?

Looks awesome. Getting an agent to actually work in production is a whole different challenge vs vibe coding automations, and this feels like it removes a lot of that headache. Excited to try this as a PM.

the 6 point gain on IFBench over the base model is pretty impressive. what's actually happening in the harness that improves instruction following that much?

The structured spec approach is a smart bet. Most agent frameworks right now ask you to wire everything imperatively, which makes it really hard to reason about what the agent is supposed to do versus what it actually does. Curious how the spec handles cases where an agent needs to adapt its behavior based on context it did not have at definition time. Is there a way to express conditional logic in the spec, or does that get pushed into tool implementations?

multi-provider routing with fallback is usually the thing that gets rebuilt from scratch on every project. either you're locked into one provider or you've added a custom routing layer on top of whatever sdk you started with. having it in the agent harness directly is the right place for it.

Does Waitlister cover opt-in for B2C marketing? I didn't bother with a waitlist for my product but now I only have the setup to generated & communicate with B2B leads. Would've been great if I captured B2C opt-in's prior to launch using a waitlist.

congrats on the relaunch. prepping my own PH launch in 2 weeks and the question i actually care about does waitlister natively segment signups by order ? trying to do honest founding-member scarcity (first 50 at price A, next 50 at price B, then standard) without faking a live counter. saw you have webhook events so worst case i wire the cutoff myself, but native would be cleaner. and fwiw the device fingerprinting on referrals is the spec i wish more tools had had to roll my own trial-fingerprint table for similar reasons.

Really like this focus. Building a waitlist before launching is still underrated. How are people typically driving traffic to their waitlist pages once they’re set up?

the AI landing page from one line is either going to be really good or really generic. curious what the output actually looks like — does it pull in real copy or is it placeholder vibes until you edit it?

Congratz on launch. Just some feedback about the page design. Basically I like that it is full screen width, but these elements get stretched out on 2560x1440p monitors so that the illustrations and text-boxes are at the opposite sides of the page making it look kind of wonky:

Really dig the approach here. Going live in 10 minutes with referrals and email automation baked in is exactly what most early-stage founders need. Curious how the referral system tends to perform compared to plain signup flows?

Waitlists sound great on paper, but a lot of them end up being just a list of emails that never really convert

The hard part is what happens after someone signs up

Would be interesting to hear how this performs when it comes to actual users, not just signups

Hey Product Hunt 👋

I built VIDEOAI.ME because I needed it myself.

Here's the truth: video is the #1 channel right now, but most of us hate being on camera.

Founders are heads down shipping. Marketers don't want to be the face of the brand. Ecom owners don't have time to film, light, edit, reshoot.

So we all just… don't post. And we lose to the few who do.

I started making AI videos just to solve my own problem.

Then it worked too well to keep to myself:

- Grew a TikTok channel to 15k followers using videos made with this tool

- Ran winning UGC ads for top ecom and SaaS brands (the kind that actually scale)

- Use it daily for my own user acquisition

VIDEOAI.ME lets you create your own AI actor (or a cast of them), give them new looks, and turn a script into a real talking video. No camera, no studio, no awkward takes. Just type, generate, post.

If you're a founder, a marketing team of one, or an ecom owner who knows you should be doing video but keeps putting it off, this is for you. Stop being shy, start shipping videos!!!

Would love your feedback, thanks for the support of this special day 🙏

The friction you're solving is real—most teams know video converts but the production overhead keeps them from actually doing it. Curious how you're handling the uncanny valley problem with the AI actors, since that's usually where people get skeptical on first impression.

Honestly this hits a real pain point. I've tried forcing myself to record videos before and it's always a mess or just takes too long... so I end up not posting atall. This feels like something I'd actualy use consistenly, especially if the actors don't look too AI-ish. Gonna give it a try and see how it performs for ads!

What if someone wants to train an AI model on a person who didn't permit it to do that?

This project looks really realistic ! Interesting, looks really realistic!! bsaed on the videos on your landing page, it could really replace UGC.

Congrats on the launch! Spinning up an AI clone from a single selfie and shipping native-feeling videos in 70+ languages is wild. Any plans to add more control over actor expressions and pacing?

Everyone’s trying to solve the “making videos easier” part
But feels like the harder problem is whether people actually watch and trust AI-generated content long enough to convert

A lot of content gets produced, not much of it sticks @grsl_fr

yo this is unreal, can't wait to build my cat feeder with atech🚀

Genuinely unique idea. I've spent way too many late nights hunched over Arduino boards, scrubbing through bad YouTube tutorials trying to build stuff that should not be that hard to build. Looking back in a few years this idea is going to seem sooooo obvious - hardware shouldn't be harder than vibe coding. Loved the vision so much I just vibed my first hardware on the site. Can't wait for the kit to show up and test the upper limits?! I wanna vibe build freaking rockets 😅

Woohoo, congrats on the launch! 👏👏👏

This is such an amazing way to bring the playfulness of LLM into the world of hardware design!
Good luck with the launch!

This is the first time hardware has actually felt vibe‑codable to me. As a software + real estate guy, I usually avoid soldering and firmware entirely. Very excited to follow this along!

The framing of hardware never getting the abstraction layers software did really lands. Curious about the boundary you draw between what the chat layer generates versus what is baked in at the module level. When I add a new sensor module, does the model see the schema and rewrite firmware on the fly, or is it more declarative than that?

Snap-together hardware with auto-generated firmware is a fresh take — most "easy electronics" stops at a USB cable to your laptop. The bit I'd want pressure-tested is the firmware-by-prompt loop. When the generated code doesn't quite work on a specific module, can you inspect and edit it, or do you reroll the whole prompt? And how does it handle timing-sensitive logic where AI-generated code can quietly miss interrupt windows?

The idea of making hardware feel like LEGO is really interesting. But feels like the real challenge starts when things don’t work as expected.

Debugging in hardware is usually where most of the friction is. Curious how you’re thinking about that part?

This is a really interesting idea, the key will be how do you communicate it in simple English!

Are you targetting the people that already know they have this problem or are you trying to generate awareness?

For example I just tried to use the query on the website front page and I got confused by the first card that popped up talking about how many ports, also I cannot seem to get past this screen?

Hybrid deterministic modules with generative system logic is a sharp split. How do you version module firmware so an LLM regen does not break existing setups?

Congrats on your launch! Real-time crowd levels and WiFi quality in one place is something that is super necessary to know when picking the right coffee shop to do your work at. As someone who works half remote and basically lives out of coffee shops, the amount of times I've shown up somewhere packed with no outlets is too many to count. The check-in feature is a great touch for keeping the data actually accurate. Are there any plans to let users filter by noise level, or vibe, like quiet vs. lively spots?

I work from home and I started going to different coffee shops to work and be around people, but I quickly realized that there are a lot of cafes that don't have plugs, good wifi or good coffee so it was always a toss-up. As a result I ended up building a tool where people can get better information about cafes before going there. Then I added a new feature where people can check-in and tell others that they are there and how busy it is and that became brefinder.io.

This is the first product I build, and I am hoping to release it as a mobile app soon but would love any feedback any of you have for me. I also have some good social feature updates coming soon and will have more info on cafes available!!!

The wifi-and-plug data is genuinely the gap nobody fills — Google Maps reviews are useless for "can I actually get four hours done here." Cold-start is the part I'd worry about though. In a city where Brew Finder is new, how do the first few real-time signals get bootstrapped before there's enough check-in volume to be trusted? And does stale data decay fast — if someone checked in three hours ago saying it was empty, does that still show up at peak hour?

Tried it out but keep hitting the ‘Location information is unavailable’ message even with access turned on.

As a dev, curious if this is built on Google Maps API with restaurant/cafe tags or using a different architecture.

finding good spots to work is always harder than it should be

but feels like the tricky part here is keeping the data actually up to date in real time

if that works well, this becomes super useful

Genuinely useful.

Got all the signals that matter when I think "can I actually work

here"?

• Wifi reliability

• Plug count

• Noise level

These are never on Google reviews!

If you are crowdsourcing those three, that is the moat!

Love this idea. How oriented is this to customer experience versus coffee quality?

I want this for Lisbon!!

As a person who works from home, I'd love this to become a bit more social :D

Does it work for places outside of massive urban centers, and for balkan countries?

Hey Product Hunt!

I built GitBar because I was tired of juggling PRs across GitHub, GitLab, self-hosted GitLab, and Azure. Five browser tabs, a Slack reminder, and I was still missing reviews.

GitBar lives in your menubar. A number tells you how many PRs need you. Click it and you get a list split into three tabs: Mine (PRs you authored), Review (PRs waiting on you), and All.

Key features:

• Multi-platform. GitHub, GitLab (cloud and self-hosted), Azure DevOps.

• Multi-account. Personal and work, side by side.

• Mine / Review / All tabs. With filters you can tune in settings.

• Status badges. Approved, draft, conflicts.

• CI status on the card. So you know when a PR is ready to review.

• Compact view. Fit more PRs on screen when you're juggling many at once.

• Hide PRs. Right-click to hide them from Mine & Review.

• Live menubar count. Glance, don't click.

• GitHub Issues. For the work that isn't code review.

• Native notifications, launch at login.

Today I'm shipping 2.0. New Mine / Review / All tabs, redesigned cards and onboarding, new app icon, GitHub Issues, and it's running on React Native's new architecture, so it's noticeably snappier than v1.

Free on the Mac App Store. Would love to hear what's missing.

This solves a real pain point. When I was running engineering at a company that grew from 15 to 120 engineers, PR review latency became our silent killer - not because people didn't want to review, but because the notification was buried in a browser tab they forgot about or a Slack channel with 200 unread messages. Having it in the menubar with that Mine/Review/All split is exactly right because the mental model matches how engineers actually think about PRs. The multi-platform support is clutch too - we had teams split across GitHub and self-hosted GitLab and the context-switching between them was brutal.

CI status on the card is the quiet feature in this launch. the loop without it: open browser, check if CI passed, close tab, go back to what you were doing. repeat. having that in the menubar removes a small context-switch that happens more times a day than i'd like to admit

Love this. Makes daily dev work feel lighter already.

This is super cool, I've built a tui to handle this at git-switchboard.com, but having it right in the native menubar is a really nice addendum

Having every PR consolidated across providers in one menubar is the kind of thing I keep meaning to build for myself and never finish. Curious how you handle review state synchronization when reviewers are working in the actual web UIs in parallel. Do you poll on focus, or is there something smarter going on under the hood?

Hi everyone!

I like @Odyssey’s framing of this as pretrained physical intelligence.

The model learns how scenes evolve, how objects move, how interactions play out, and how the world holds together over time. Odyssey-2 Max is still early, but this direction makes the robotics, gaming, simulation, and interactive systems angle much easier to understand.

Can someone else submit newspapers?

Hey Product Hunt! 👋

I'm excited to share SNEWPapers — the world’s first AI-powered historical newspaper archive. We’ve read and organized 6 million+ stories from 250 years of American newspapers (1730s–1960s) so you can finally explore history by meaning, not just broken keywords.

Maybe the biggest news since sliced bread for digital humanities, historians, researchers, genealogists?

I built this after trying to research references in The Fourth Turning. Traditional archives dumped me into faded page scans with terrible search. So I created my own.

The result: clean, summarized articles and nearly perfect full-text OCR extractions + The Sleuth (your personal AI research assistant), smart categorization (24 categories / 1,000+ sub-categories), Collections for sharing, and a fun Today in History daily feed.

Quick start (10 minutes): → Tutorials

A few things I’d love your thoughts on:

• Today in History — Would you actually open this daily?

• Search + Sleuth — How useful is semantic search and the AI assistant for your research?

• Collections — Would you use/share public collections?

Pricing: 7-day free trial. I priced it ~50% below traditional archives because we actually deliver usable, intelligent access. Product Hunt special: Use PRODUCTHUNT20 for 20% off any plan (valid until May 8).

Huge technical journey. I had to figure out how to acquire, store and process nearly a million high-resolution newspaper images, build custom multi-modal systems to detect and segment articles, massively improve OCR on centuries old ink, train models to understand newspaper layout and context, run prompt engineering at scale, balance cost vs quality with LLMs and vLLMs, build semantic and agentic search infrastructure that actually works on millions of documents, and scale a cost-effective GPU fleet.

Some “AWS-ish” stats so far:

• 115,000+ GPU GB-hours (OCR / Layouts)

• 26,000+ Lambda GB-hours moving data around

• 44.7 billion LLM/vLLM tokens processed

• 7 months of 80+ hour work weeks (organic neural network compute)

Would love your honest feedback and discoveries you make in the archive! 🫡 (here or hello@snewpapers.com)

Honestly, this is quite cool!

Do you plan to expand the newspaper libraries to other countries?

Cool!
What APIs did you use to scrap all the Newspaper archive?

Incredible scale!
You mentioned training the model to handle degraded paper and faded ink. Google famously used recaptcha v1 for the same problem, having millions of users unknowingly label words from old NYT archives. How have you coped this issue?

You're onto something real here. I managed engineering teams for years and watched Jira boards become graveyards - hundreds of tickets nobody looked at, zero sense of momentum, and sprint retros where everyone just shrugged. The core problem isn't the tool, it's that task boards show work as an infinite list rather than a journey with visible progress. Reframing tasks as missions on a shared map is a clever way to give teams spatial awareness of where they are and where they're heading. Curious whether you've seen this change how teams handle scope creep - when progress is visual like this, does it make it harder for stakeholders to quietly add tickets without anyone noticing?

Week one with a tool like this is easy. The map view can be really motivating. But I'd want to know what week eight looks like? When the skill cards feel routine and the tree of progress is just... there. Does PlayJoob still carry teams through a rough sprint?

Congrats on the launch!

Hi everyone, thanks for checking Wafaa.io.

Building Wafaa began with a simple idea: creating a secure contract should be as easy as creating a WhatsApp group.

For many people, contracts are low-frequency but high-urgency. When you need one, you tend to need one immediately—but most of the existing tools feel overly complex or like they're built for legal teams and large enterprises, not everyday users.

At the same time, digital commerce has a growing problem: fraud and disputes are becoming more common, especially when agreements are informal or poorly documented.

We built Wafaa to address both.

In short, the platform simplifies how agreements are created, while adding structure, traceability, and verification. This helps both parties know exactly what was agreed to, and can prove it if needed.

Instead of replacing lawyers or overcomplicating the process, Wafaa focuses on making agreements clear, structured, and reliable from the start—and providing users with a centralized hub to keep all contracts organized in one secure location.

We're still early and have plenty of room for improvement — would genuinely value your feedback.

the postioning toward non-legal users is interesting. A lot of people avoid contracts just because the tools feel intimidating.

The built-in fraud prevention angle is interesting. How do you actually verify counterparty identity at signing without making the flow feel like a bank KYC?

All the best Yousif.

I could see a world where teams publish context packs for their APIs so that anyone building integrations with them gets better AI results out of the box. Are you seeing that kind of use case emerge yet, or is it mostly individual teams packaging their own internal context?

This hits a real pain point, especially losing context between sessions and having to rebuild the same thinking over and over. I like the shift from just prompts to capturing the actual workflow and decisions behind the work, that part feels more useful long term. I’ve run into this a lot switching tools mid-task, so this resonates. I just upvoted it, nice work shipping this.

P2P knowledge sharing is genuinely interesting because the centralized model has so many failure points — the platform changes, goes down, or starts optimizing for things you don't care about. Local-first makes real sense as a counter to that. Curious how you're thinking about discovery though: how do people find relevant knowledge from others on the network without some kind of central index pulling it together?

Hi Product Hunt community,

The other day, I caught myself thinking about how outdated B2B sales and marketing still are: inefficient, expensive, and full of intermediaries just to reach your own customers.

So I asked a simple question: what if supply and demand could be matched in real time using algorithms and AI agents?

Why spend time on repetitive outreach or burn budgets on rising ad costs, when AI agents could handle discovery, qualification, and even negotiation, while humans step in only at the final stage?

That’s how White Rabbit was born: the first “asocial” B2B network built for AI agents. Think of it as a stock exchange for B2B products and services, or a version of LinkedIn where posting and conversations are handled entirely by AI.

This isn’t just another commercial project for me. It’s a mission to level the playing field between large and small companies, giving everyone equal access to demand.

If that resonates, join me on this journey 🚀
Davit

P. S. Feel free to join our Discord as well!

congrats on the launch Davit - great work!

Just a question on AI agents - does it only work well if there is a big pool of AI agents within White Rabbit or you'd also fetch / try to interact with public AI agents outside of the tool? and can I create multiple AI agents at the same time as I have different service offerings etc.

WOW congrats for launching such a product. I haven't seen this category of product before!

Thanks for the encouragement. The positioning came from watching how much time our early users spent manually sorting through noise on traditional networks. Flipping it to let AI agents do the heavy lifting felt like the natural next step.

congrats on launch! b2b sales really does feel broken in many places. Curious how you’ll solve trust between AI agents – that feels like the key part here

Congrats on the launch Davit!

Wishing you a successful launch! :)

Love the "asocial professional network" framin, flipping LinkedIn for AI agents instead of humans is one of the boldest positioning lines I've seen on PH this week. Rooting for where this goes!

Congratulations. And happy product launch. @davitausberlin

I'm Tae — solo founder, also solo marketer, also solo support team. You know the drill.

Mirr started from a selfish problem: I was spending 3 hours a day making carousels, scripts, and captions just to stay "consistent" on socials. Meanwhile the actual product wasn't getting built.

A social media agency quoted me $3K/month. Absolutely not.

So I built the thing I wanted — an AI team that takes one idea and turns it into a week of content across Instagram, Threads, X, and LinkedIn. In my voice. Scheduled. Published. Done.

A few things I'm proud of:
→ One prompt → carousel + short-form video + blog + captions (not just captions like 90% of AI tools)
→ Actually learns your voice from your past posts
→ Scheduling + comment inbox + analytics in one tab, not five

🎁 PH exclusive: code PRODUCTHUNT = 40% off your first month.

Honest question for you: what's the one part of social media you'd pay anything to never do again? I'm reading every comment today.

A service with massive growth potential. The essential AI for SNS marketing that early-stage startups need most. An automated traffic-hunting machine.

Been using Mirr and genuinely love how much time it saves. As a founder, social always felt like something I should do but kept pushing off. Mirr makes it way easier to turn one idea into consistent content without juggling five different tools. Huge congrats on the launch!

been waiting for something like this. as a founder, managing social media has always been such a pain. congrats on the launch, mirr! gonna try it out this week

Cool idea, I’d love to give it a try. How does it work for Reddit, does it mirror posts there too?

Not gonna lie, most “AI content” tools fall apart at consistency.
Week 1 looks great, week 3 it’s already dead.
Curious how you’re solving that long-term

Hey Product Hunt 👋

I'm Joseph, co-founder of Anthum. We built this after watching founders burn $10k+ on agencies, only to get one creative direction they hated and couldn't change.

The real problem wasn't the content. It was being locked into one vision.

Anthum was born to fix that by connecting brands with hundreds of professional video creators who compete to make your ads. Run a contest and only pick your favorite concepts, hire a creator 1-to-1, or use our self-serve AI-generation tools to make ads in-house.

No agencies. No overhead. No AI slop. Just premium ads, made by pros, at the scale you need to find what actually works.

Would love your feedback — happy to answer any questions 🙌

Hey, I'm Naitik, Founding Engineer at Anthum.

Excited to be live on Product Hunt today.

Would love to hear your thoughts.

Been competing on Anthum for a few months and honestly it’s become my favorite thing to work on! Real brands, real briefs, and a community that actually watches your work carefully. It’s pushed me to experiment with things I probably wouldn’t have tried on my own. Pretty rare in the AI space right now, worth it.

Thrilled to be live today! We've spent months building the infrastructure to turn chaotic AI video contests into a seamless, beautiful experience for brands and creators. Would love to hear your feedback or talk shop about the build!

Hi, I'm Andriyand, one of thousands of AI creators that Anthum has. I've been joining Anthum for the past few months and participating in their contests. So far, Anthum has really helped me build my portfolio with a very supportive community.

I've been using Anthum for 4 months and I'm very happy with it. Very much so. I'm a simple user, with no financial or personal connection to the site's owners. I consider myself an expert in AI sites because I have accounts everywhere, and Anthum is the one I find most useful. I use Anthum Studio almost daily and it's a great tool. I've earned credits very easily just by judging contest videos and placing in a couple of them. Their Discord is a place full of creative and friendly people, and I've attended some really interesting webinars.

I’ve been creating on Anthum since the early days, and I keep coming back because of the quality of the creators on the platform. It's an inspiring group! In every contest, someone submits work that teaches me something new and sharpens my approach for the next one. Great work, guys, and huge congrats on the PH launch!

Hey I'm Vitaly, I have a really positive experience with Anthum AI. What I like most is that it gives creators a real opportunity to showcase their skills in AI ad creation while working with actual brand briefs.

The platform is well-structured, with consistent contests, solid prize pools, and clear creative directions. It feels motivating because there is always a new challenge to work on, which helps improve both creativity and storytelling skills.

I joined Anthum a few months ago and am really happy I did. I enjoy the contests, the workshops, and the community. I’ve even fully transitioned to using their built in Studio for creators and it was a decision I am completely happy with. I fully support Anthum and want them to succeed beyond all they could have imagined and I know they will.

I gotta say, using Anthum as a creator is really good experience for me. I liked the peer judging feature a lot because it actually helped my first video get noticed and gave me useful feedback. The platform is easy to navigate, so I didn’t struggle to find my way around or create content. Everything feels simple and smooth while using it. I also feel like Anthum still has more to offer as time goes on, which makes me excited to keep using it.

Anthum truly is one of the greatest platform out there to both create with AI and get work directly through their system connecting with agencies. I have been using their services for the past few months and was able to create some of my best work (cheap, reliable generation models, Kling 3 , Multi reference Seedance 2.0) and affordable. Their frequent contests provide a true professional experience and encourages every users to push the boundaries with AI, get better, learn and share with other members of the community. The team is extremely active, responsive and helpful. I recommend it to anyone who want to get their feet into AI filmmaking for both beginners and advanced. Here is a short clip make with anthum: https://x.com/MoussTvAI/status/2046269020949655721

Spent a few months on Anthum and it’s easily become my go-to. Real brands, real creative briefs, and a community that actually cares about the work. It’s pushed me to experiment in ways I normally wouldn’t. Honestly, that’s rare in the AI space right now—definitely worth it.

Hi project hunt

I'm Zivah

A creator under Anthum AI, I joined the platform few months ago and from my experience with them this few months, I've learnt how to upscale my skills in AI, learnt how to prompt, edit, camera shots and so many stuff I would need funds to learn, but I got it all in Anthum AI community where knowledge is shared every single time.

They most importantly help other projects and platforms to grow and gain visibility through contest and other means.

From what I can say, Anthum AI is not just a platform but a community that carries its members along, fulfilling a great future where every one benefits.

In my 3 weeks using Anthum i have became part of a network of peer reviewing AI videos, grounding market opportunities, creative passion and product vision into a very confortable workflow. Product and manufacturers requesting ads to follow branding guidelines, A new world to land as a marketing analyst graduating to use AI for specific goals in serious marketing campaigns.

I wonder, is Anthum aiming to level the playing field and give everyone a fair shot at competing with the big advertising agencies? hmm, Exciting! Can't wait to see if Anthum will bring in those major brands that we usually don't have access to.

I think it has potential, I entered a contest, and it was quite a bit of work, very intelligent system of constantly referring back to the creative brief, as you chat with the ad helper, it guided me to completion and I submitted. I would have added more bells and whistles to the ad, but they require you use their Ad maker, which is robust enough to get the tasks done. I did find the needing to buy credits to finish up the video renders to be somewhat gimmicky. Its almost like gambling, you pay $20 (how much credits i paid for ) for a chance to win, $5000? AI video is a passion for me, so i dont mind, crunching through commercial stuff once in a while. I like their discord as well, informative, they found me on kling discord which I thought was smart too.

Saya sudah menggunakan Anthum 3 bulan lamanya. Saya mendapatkan pengalaman berarti dalam mengerjakan tugas tugas yang diberikan oleh Anthum. Saya benar-benar berlatih untuk bisa menjadi juara walaupun belum dapat. AI nya lengkap, mulai dari buat image, membuat suara, hingga editing video. Yang perlu diperbaiki adalah layer di editing, entah saya yang belum bisa menggunakannya atau memang belum tersedia.

I mean, as a user, Anthum has really called up my video content creation games. From engaging in some of their recent contests where we created video content for an health brand and also one for a bed mattress production company and even now to their current video ads, for someone like me without a very heavy background in video content. And I'm looking towards exploring beyond the limit of what this platform has to offer.

Hey everyone! 👋

I'm Basil, founder of Anjiz.

I've always loved starting small businesses. But every time, I hit the same wall — tracking inventory in my head, managing orders on WhatsApp, writing expenses in a notebook, and never really knowing my actual profit. The tools out there were built for big companies, expensive, and each one only solved one problem. Nothing was built for someone like me.

In 2022, I decided to build it myself. By March 2023, I was finally making progress — then war broke out in my country. My family and I lost everything. For three years, I couldn't touch Anjiz. We left the country, started over in a new city, and I focused on surviving — finding a job, rebuilding from zero. But day after day, Anjiz stayed in my head. The dream never left.

Now, after 4 years Anjiz is live. And in just our first three days, 1,000+ businesses from 34+ countries signed up.

One app for everything: products, orders, invoices, expenses, CRM, tasks, and reports..

I'd love your feedback 💬