PH热榜 | 2026-06-02

Hey Product Hunt! 👋 I'm Anna, founder of Fundraisly.
I spent 2.5 years as an investment analyst at $600M+ AUM VC Fund, portfolio includes 10 unicorns. I reviewed thousands of pitch decks — and saw firsthand how broken fundraising is. Brilliant founders wasting months cold-emailing the wrong investors. Meanwhile, the right ones were just sitting in databases nobody knew how to use.

So I built what I wished founders had when they came to us: an AI agent that analyzes 300K+ investors and millions of deals to find exactly who's active, relevant, and likely to respond — in minutes, not months.

The results blew my own expectations:
🎯 60–70% open rates. We only reach investors who are actively investing in your space, not generic cold lists
📞 On average, founders conduct 20-40 qualified investor meetings within the first 90 days with funds actively investing in their space
💼 3k+ VC calls conducted in last 6 months with funds like a16z, Sequoia, Index Ventures
💰 $100M+ raised for founders through the platform

Fundraisly isn't a CRM or a database. It's an AI agent that does the entire investor research, outreach, and follow-up for you — so you can focus on building your company.
I'd love your feedback — especially from founders who've been through the fundraising grind. What was the most painful part for you? Happy to answer any questions! 🚀

I like that Fundraisly focuses on active and relevant investors, not just “more contacts.” That feels much more useful for founders (or at least for me).

Curious how you decide which investors are actually a good fit for a startup. Is it mostly based on past deals, current activity, stage, geography, or all of these together?

Congratulations on the launch, we were looking for something like this for our fundraising. Can I ask how you take care of the email deliverability?

A very interesting AI business app!

Good luck!

The painful part is usually not just finding investors but knowing who is actually relevant right now. I like that this focuses on active investors instead of just another large database. How do you tell if a fund is currently investing in a specific space not just historically interested?

Skeptical about quality, ngl. Used a similar service last year, won't name names. They booked 12 meetings, but 8 were with associates at funds that did not invest at the stage or check size we needed. By the third call I was burning founder time just to hear "too early for us." How are you screening for stage fit beyond what a fund says on its website?

Do follow-ups go out automatically? That's where a lot of outreach starts feeling robotic.

Hey there

Are you in any way incentivized in a successful fundraising by your customers?

I mean, is your business model is “pay for our service” or rather “pay for your result”?

Super valuable tool to help match make when looking for the "right" money in your deal.

what happens to a founder's reputation with investors if the outreach volume is high and the targeting is off. investor networks are small and word travels. a founder who sends 200 poorly targeted cold emails through an AI agent can do real damage to their chances before they ever get on a call. how are you thinking about the downside risk of scale outreach in a community where relationships and signal matter more than volume

Fundraising is such a grind. Love the focus on warm paths here. Congrats on the launch!

Amazing product. Good luck!

Congrats on the launch!

Really interesting - curious how you handle the warm path mapping when someone's network is mostly in a different industry. I'm coming from institutional finance/tax consulting, so my warm connections are mostly family offices and healthcare executives, not traditional tech VCs. Does the system weight domain-relevant investors even when they're not traditional tech VCs? Congrats on the launch

Quite an interesting product. I happed to search for a similar one today morning and end up doing a deep research. Definitely give it a go.

Good luck on your launch! Looks awesome 🔥

Congrats on the launch

Looks like an amazing products !
do you also serve other industries thank the tech space ? Hospitality for example?

would be intereted then !
Best

Congrats on the launch! How does this compare to Carta's investor matching?

Congrats on the launch!

Market is rough right now. Wondering if these numbers actually hold up.

The product is interesting and looks impressive. I couldn't find any information about pricing or plans )

Congrats! Useful idea!

Good luck with the launch. Is this US-only, or do you cover Europe and Asia too?

Congrats on the launch! What if my round isn't quite ready yet, but I want to map the market?

If I use you for Seed, can I come back for Series A with the same data?

Wishing you a strong launch. How much of my time does onboarding actually take?

Congrats Anna and team. Fundraising tooling definitely needs a rethink.

Hope the product keeps growing. Can outreach work in languages other than English?

I keep coming back to this line from @zhen_han : “Vokal is the collaboration space for 10x teammates and their AI agents.”

Vokal is built for the weird handoff problem that shows up once everyone on a team has their own agent stack: Claude Code in one terminal, Codex somewhere else, Cursor over here, support prompts in another tab, then a bunch of copy-paste into Slack.

The product treats agents less like private sidekicks and more like teammates with roles, owners, permissions, memory, and a unified event log. So, a 10x teammate is a human who works with a crew of agent helpers.

And the flow looks like this:

Humans set goals → agents do work → humans review ↺

That feels like the right frame: not “another AI chat app,” but infrastructure for the awkward middle stage where startups are already working with agents… just not together. Yet.

Hey Product Hunt 👋

I’m Zhen, founder of Vokal. Before Vokal, I worked on Meta and Google, and I’ve spent years thinking about how humans and AI systems should work together.

Vokal is a collaboration space for 10x teammates and their AI agents.

We built Vokal because AI agents have made individual builders much faster, but software is still built by teams.

Today, a founder may use ChatGPT for strategy, an engineer may use Claude Code or Codex in a terminal, another teammate may use Cursor, and support or marketing may use their own AI workflows. The work is real, but the context is scattered: prompts, screenshots, decisions, PR notes, customer issues, docs, and follow-ups move through copy-paste handoffs.

Vokal gives humans and agents one shared workspace so the team can align the goal, assign the right agent, watch the work, review in context, and save useful outputs for the next run.

Here’s how it works:

1. Bring teammates and agents into one shared workspace.

2. Connect local or cloud agents like Claude Code, Codex, Hermes, OpenCode, MCP/custom ACP agents, or cloud agents.

3. Give each agent a name, role, owner, permissions, app access, and memory scope.

4. Run work in channels with tasks, docs, routines, Memory, and Knowledge Base attached.

5. Nudge the work in context and save useful outputs so the next teammate or agent can start from what the team already learned.

Why startups use Vokal:

• Make agent work multiplayer: agents work where teammates can see goals, blockers, outputs, and decisions.

• Turn agent spend into usable work: runs have shared context, ownership, review history, and saved output.

• Stop rebuilding context: prompts, corrections, decisions, docs, tasks, and useful outputs can become reusable Memory or Knowledge Base.

• Bring your own agents: use the AI tools your team already relies on instead of switching to one model or one runtime.

• Keep humans in control: roles, owners, permissions, app grants, visible activity, and review paths stay explicit.

Most AI tools make one person faster. Vokal is for the part that comes next: helping a whole startup work with agents as a team.

🎁 For Product Hunt, use code 10XTEAMMATES to get 1 month free.

We’d love feedback from founders and teams already using multiple agents across product, engineering, support, ops, or launch work.

The 'turn agent spend into usable work' line really resonates. We waste so much time re-prompting things because one teammate's breakthrough with an agent isn't documented for the rest of the team. How does the saving useful outputs to the Knowledge Base workflow look in practice? Is it manual or AI-assisted?

the copy-paste handoff between slack and whatever agent you're running is so real. half my team's context gets lost in that gap. one workspace where the agents and humans are in the same thread makes way more sense than the screenshot-in-slack workflow we're doing now

Congratulations

Congrats on the launch! Curious what happens when two agents disagree on the same task does Vokal flag the conflict somehow or just pick one of the outputs?

I can see it work well for non-technical collaborators & AI users, but for engineers, why is it better than a well set repository with skills, subagents, or other assisting markdowns? Would love to know more

The emphasis on visible work is important. If agents are doing meaningful tasks, teammates need goals, blockers, outputs, and review history.

Strong launch. Vokal feels like an operating layer for teams moving from “we use AI tools” to “agents are part of how work gets done.”

The unified event log is interesting. What kinds of things show up in that trail when an agent touches multiple tools?

The 'agents as teammates with roles and memory' framing is sharp. Most tools treat agents as personal sidekicks, but the real friction is when multiple people on a team each have their own stack and context gets lost in Slack paste. Curious how you handle permission boundaries — can an agent access shared memory across different user accounts, or is memory scoped per owner?

The handoff between agent stacks is a real pain. How does Vokal decide which agent owns context when two are working on overlaping tasks? Is there a permission layer per repo, or just per workspace?

What does human review look like before an agent output ships?

the review step in that flow is where most teams actually break down. everyone can set goals and agents can do work, but 'humans review' requires a skillset most teams havent developed yet — knowing what to check, how deeply to verify, and when to trust vs question the output. curious how Vokal handles that evaluation layer.

Is this more like a peer programming where coworkers can prompt / work with AI agent within the same context ?

Congrats on the launch! I’ve seen a few projects like these, and my experience tells me that indeed, keeping team in sync becomes a bottleneck in this fast AI dev tooling world.

How does your tool approach integration with team’s agents, for instance Claude/Code? Does it replace the «brain» of that tools with its own, or integrates it via MCP/other means, or both?

How granular are the app permissions? I’d want agents to access the right tools without giving them the whole company.

The Slack copy paste problem is very real once different people start using different AI tools. I like the idea of agents having roles and owners instead of everyone keeping their own private workflow. The useful part for teams might be less abt adding another AI tool and more abt making the work visible enough for others to review and continue. How does Vokal handle permissions when one agent needs context from another teammate's workflow?

Love the idea of giving AI agents a shared workspace instead of having context scattered across chats, docs, and screenshots. Congrats on the launch!

Does Vokal read all company data by default, or can teams scope what each agent sees?

How would a support team use this when a customer issue needs to become an engineering task?

How does Memory / Knowledge Base work in practice? Is it more like saved prompts, team decisions, or both?

Hi everyone, I’m Namanyay from Gigacatalyst (https://gigacatalyst.com/). Gigacatalyst allows sales, CS, and users to build one-off features, so your SaaS can support long-tail customer workflows and engineers aren’t pulled away from the roadmap.

When you sell software to large businesses, you realize that each customer needs their own workflow and features. Traditionally, this either means long engineering roadmaps or the customers end up using workarounds.

But what if everyone could build their critical missing features just by talking to an AI? That’s what we do at Gigacatalyst. We provide an AI customization layer for your customers, CS team, and sales team to build these missing critical workflows without needing any engineers at all. Think Lovable, but built on top of YOUR platform.

We connect to your product's APIs, learn your data model and design system, and let non-technical users build governed apps via natural language - inside your product, under your brand.

These folks are scary good. I'm split between the philosophy and the product: as a user I love the idea of emacs-like software where computers grow into human needs, but the more relevant part for companies is that you simply cannot plan out everything a user needs to see/do, and you want to allow them to solve their problems as they see fit. And you can't do that without (something like? but I haven't seen anything remotely like) @Gigacatalyst

Super cool concept - congrats on the launch!

Congrats on the launch @namanyayg would love to give it a shot!

The positioning is way too great and launch video is awesome love it

This is such a killer product. When I worked at Datadog, we actually built internal tools for this!

Been using Giga in Scalio to provide our customers with instant AI website generation. The response has been great, and it's saved us months of in-house development. Instead of building all the infrastructure ourselves, we were able to focus on our core product and get to market much faster. Big fan of what the team is building.

Giving non-technical users technical powers is super powerful... especially in a world where customization is cheaper than ever. My only concern is that sales/CS often don't have a good "radar" for what should be built (even if it can be built).

Congrats on the launch Namanyay!

The idea of letting non-engineers build features is genuinely compelling. But what happens when your API changes six months later and a customer's custom workflow breaks? Who owns that - you, the customer, or Gigacatalyst?

Congrats on the launch!

the customer retention angle makes sense but the risk is that custom implementations per customer make your product harder to maintain over time not easier. every bespoke workflow is technical debt someone has to own eventually. curious whether the customizations are sandboxed per customer or if they can affect the core product

can i plug custom MCPs here?

This is so cool! What are the type of guardrails you are adding to protect from misuse?

Really good product that gives tons of flexibility to the team! Congrats on the launch

Congrats on the launch!

Congrats for your launch, The workflow-first approach makes a lot of sense

Really interesting approach to customer-driven product building. We used to think of Saas as public transport that gets you 80% there. Seems more like that transforming to Uber

How much API documentation or setup is needed for building features with Gigacatalyst?

Let's go guys !!

Hi Product Hunt! 👋

I’m Franklyn Wang, founder and CEO of Liquid.

Before starting Liquid, I led AI for Macro at Two Sigma, working at the intersection of machine learning and markets. Before that, I was Harvard’s top-ranked math student twice.

One thing became clear throughout my career:

The best investors don’t win because they’re lucky.

They win because they have better information, process it faster, and execute with more precision.

For decades, that edge has mostly belonged to professionals.

Professional investors have had:

• Better data

• Faster research workflows

• More sophisticated models

• Real-time market intelligence

• Direct execution infrastructure

Everyone else was guessing, or stitching together fragmented workflows.

That’s what we’re changing with Co-Invest: the first way to trade inside AI assistants like Claude.

Co-Invest gives everyone access to the kind of investing workflow that used to be reserved for professionals:

• Superior information, including on-chain data like positioning

• Superior data processing, powered by frontier AI

• Superior execution through Liquid, with some of the lowest fees available

Instead of just reading about markets, you can ask questions, analyze opportunities, build conviction, and act — all through conversation.

Co-Invest lets you:

• Access real-time market and on-chain intelligence

• Construct portfolios according to any criteria, from risk profile to values

• Process complex data faster with AI

• Execute trades directly from chat

• Access 500+ markets, including crypto, equities, FX, commodities, and prediction markets

• Trade 24/7

Why we built it:

In high school, I was struck by the idea of rational ignorance: because humans have limited time and limited ability to process information, it can be rational not to know things. Sometimes, the cost of becoming informed is higher than the benefit.

I hated that idea.

As a math-obsessed kid who craved perfect rationality, the notion that ignorance could be produced by the system — not just by the individual — felt like surrender.

As I grew older, I saw how expensive that ignorance was.

Finance, at its root, is about asymmetric information. One side of every transaction knows something the other side does not.

An influencer pumps a stock to followers, then exits at the top. A scammer convinces a retiree to wire away her life savings. A brokerage auctions off the right to trade against your order. A market maker pays for the right to see your order before it fills.

Some of this is illegal. Some of it is legal. But every example has the same shape:

Wealth flowing from the less sophisticated to the more sophisticated.

Modern finance is very good at making that happen.

Co-Invest is our attempt to change the equation by giving everyone access to superior intelligence, better information, and better execution.

Who it’s for:

• Investors who want better research and execution

• Traders managing opportunities across multiple markets

• AI-native users who already spend their day inside assistants like Claude

• Anyone who wants access to a more professional-grade investing workflow

Our vision:

At Liquid, we’re building toward an AI-native financial system accessible to everyone, everywhere.

Our first product, the Liquid mobile app, gives people access to every market. Co-Invest pairs universal market access with universal intelligence.

That combination — universal markets plus universal intelligence — is incredibly powerful, and we’re just scratching the surface it.

We believe the next generation of investors won’t be limited by who has the biggest team, the most terminals, or the most infrastructure.

They’ll be empowered by AI.

Co-Invest is our first step toward making professional-grade investing capabilities available to everyone.

If you try it, I’d love your feedback:

• Which markets, brokers, or exchanges should we support next?

• What’s the biggest bottleneck in your investing workflow today: information, conviction, execution, or risk management?

Thank you for checking out Co-Invest and supporting our launch.

Looking forward to answering your questions in the comments.

— Franklyn Wang
Founder & CEO, Liquid

Franklyn, congrats on the launch!

Love how you brought the terminal into the chat interface! I'm curious about trade confirmations though.

Are you use a UI modal before execution to prevent accidental trades, or do you handle risk entirely with plain-language constraints in the chat?

since you are offering 500+ markets across crypto, equities, FX, commodities and prediction markets, which asset class exposed the biggest technical challenge in terms of data normalization is it the fragmentation of on chain data across different blockchains or the regulatory complexity around executing equities inside a conversational AI?

Interesting

500 markets across crypto, equities, FX, and commodities through a single interface raises some obvious questions about regulatory coverage. equities and FX have very different compliance requirements depending on the user's jurisdiction and 'fund in-chat' for securities trading is the kind of phrase that makes compliance teams nervous. curious how you're handling the regulatory layer across different asset classes and user locations

Strong concept. Security question: if my conversation with Claude or ChatGPT contains trade instructions and account details what's the data retention policy? Is the trade context stored on Anthropic or OpenAI infrastructure on your servers or is it ephemeral per session? For institutional users especially that distinction matters quite a bit.

Congrats on shipping this. I'm trying to think through the liquidity side for equities and FX specifically, are you acting as the broker dealer directly, or routing through a prime broker or third party liquidity provider? That affects execution quality, margin requirements, and ultimately how competitive your pricing can be at scale.

The AI handles support with full context feature is one of the most interesting parts of this product to me. Does that mean the support agent has read access to my full trade history and positions or just the current session? And is that handled by the same model or a separate system? The context boundary matters a lot for how much I'd trust it.

Built a Kalshi prediction market trading agent and Alpaca equity agent from scratch - so curious about the execution layer here. Alpaca gives developers raw API access to build their own logic. Co-Invest seems to be positioning as the consumer-facing version of that same idea. How are you thinking about the power user who wants to automate recurring strategies versus the casual trader who just wants to ask Claude what to buy? Those feel like pretty different products

Curious about the prediction markets integration specifically are you routing through an existing platform like Polymarket or Manifold, or have you built your own liquidity layer? Prediction market liquidity can be thin on non headline events, so I'd love to understand how you're handling slippage and order size limits there.

The plain language execution is compelling. I'm wondering how the AI handles ambiguous instructions for example, if I say 'buy some ETH when it dips' does it ask for clarification, set a conditional order, or wait for explicit confirmation before doing anything? Understanding where the guardrails are would help build confidence in using it for larger positions.

Congrats on the launch. Trying to understand the business model are you taking a spread on trades, charging a flat subscription, or is it fee per execution? With 500+ markets across so many asset classes, I imagine the revenue structure varies a lot by product type. Would be great to have more transparency on that upfront.

Love the on chain signal layer. Quick technical question: are the wallet activity and order flow signals pulled in real time or is there a lag? For fast moving markets like perps, even a 30-second delay on funding rate data can meaningfully change the trade case. Would love to know the data freshness story.

Really impressive scope here curious about the security model. When a user funds in chat and places a live trade, where are the keys or credentials held? Is execution going through a custodial wallet, a connected non-custodial wallet, or something else entirely? That's the piece I'd want to understand before putting real capital in.

You said finance at its root is about asymmetric information where one side knows something the other doesn't what specific feature or piece of data inside Co-Invest today gives a retail user an informational advantage that even many small professional funds still lack access to?

trading real money from a chat prompt is one of those things that sounds amazing until you think about what happens when you word something wrong. the confirmation step before execution matters a lot here

How does Co-Invest handle position sizing, portfolio correlation and risk management across multiple markets can i ask it to never allocate more than 3% of my total capital to any single prediction market outcome, and if two of my positions are correlated above 0.7, warn me before I add a third and have it enforce those rules conversationally?

for an AI native user who already spends their entire day inside Claude for work, research and coding, how does Co Invest change the marginal cost of checking a market hypothesis do you see users moving from I will look at that later to acting on ideas immediately because the friction from curiosity to execution has dropped so dramatically?

What safety precautions do you have, like confirmation steps and limits, so a bad prompt can’t accidentally execute a huge trade?

Hey Product Hunt, I'm Drew, co-founder of Brief. Huge thanks to @chrismessina for hunting us!

Brief is a teammate that knows your product cold: why every decision got made, what you ruled out, and where you're headed. Ask it in Slack or chat. Your coding agents ask it too, over MCP and CLI.

The problem

Every time you spin up a new coding agent, you re-explain six months of decisions. Why the schema looks the way it does. Which approach you already ruled out. What the customer actually asked for. With no context, the agent confidently ships the wrong thing and you waste time.

Give that same agent access to Brief and it follows your team's decisions 95% of the time, up from 46% on the codebase alone. In our benchmark, 8 of 8 tasks came back merge-ready versus 2 of 8 without Brief, at 68% lower cost per shipped task.

How it works

• Point Brief at GitHub and your task manager (Linear, Jira, etc.)

• 20+ agents catalog your decisions, encode strategy, research users, and map competitors

• Spin up a new coding agent and Brief onboards it for you, pulling in just the context that matters

• Wire it in with npm i -g @briefhq/cli then brief init, or connect any agent over MCP at https://app.briefhq.ai/mcp

Nothing new to learn. No migration. Brief reads the work you're already doing and keeps the ship pointed in the right direction.

Who it's for

For indie hackers and early teams, Brief is your product strategy partner. For scale-ups, your executor and decision keeper. For AI-pilled enterprises, a transformation force multiplier.

🎁 For Product Hunt

3 Months Free, plus Brief will guide you through building a killer Product Hunt launch strategy. Expires at midnight June 10th.

👉 Get started at briefhq.ai. Point it at GitHub and watch the agents catalog your last 6 months of decisions in minutes.

Would love your feedback and your roasts. We're in the comments all day. 🙌

AI Ships. Brief Navigates.

- Drew

Decision traces are underrated, this is so cool!

We've found that one of the hardest parts of building agent products is figuring out whether a failure is a model problem, a workflow problem, or simply the wrong target user.

How are teams using Brief in practice today? More for understanding user behavior or for iterating on agent workflows themselves?

the problem Brief is solving is real but i'm curious about the adoption curve. the teams who would benefit most from this are the ones moving fastest and documenting least, which means they're also the ones least likely to build a new habit around capturing decisions. how are you thinking about getting context into Brief without creating a documentation tax that slows down the teams you're trying to help

Tried Brief, and the part I loved is that it doesn’t treat product context as just another doc or ticket. I've seen a lot of agent failures, as the agent can read the code, but it doesn’t know why certain decisions were made or why some paths were already ruled out. Having that decision history available instead of starting from scratch every time feels really useful. Congrats on the launch !!

Congrats on the launch @drewdil and the @Brief team. Super cool - the context awareness especially for agent-to-agent collab makes total sense.

The "re-explaining six months of decisions" line hit me right in the chest. I once watched an agent confidently refactor an entire auth module based on best practices, completely ignoring the three-week Slack thread where we'd explicitly ruled out that approach for compliance reasons. It was technically perfect and strategically useless.

Having something that sits between the codebase and the agent to preserve that institutional memory feels like the missing piece everyone's been working around with increasingly elaborate system prompts. This is exactly what teams burning tokens on rework need right now. 🙌 @drewdil

much needed @drewdil @ryanmindigo @kasyap_varanasi_ ! upvoted :)

Question: how exactly is it different from just giving agent access to Jira/github? Agent can get PR list, connect with different work items and figure out the direction we are moving. How does Brief make this process different? Thank you :)

The decision history part is the most useful piece here for me. A lot of agent mistakes happen because the code is visible but the reasoning behind the code is not. If Brief can show an agent why something was built a certain way or why an option was already ruled out that could save a lot of repeated work. How do you decide which past decisions are important enough to bring into a new coding task?

The problem you're solving — coding agents shipping the wrong thing because they lack institutional context — maps directly onto a discovery problem too: knowing what users are actually saying about products like yours across Reddit, Hacker News, and niche forums before you encode that signal into Brief. MentionFox's lead and feedback tracker surfaces text posts and audio mentions in videos where developers complain about context loss in AI coding workflows, giving you raw user language to feed into Brief's decision catalog. Check https://mentionfox.com — in 30 seconds you can see live mentions of your exact problem space and turn community frustration into validated product direction.

Great! Will it be accessible via Setapp as well?

@protsenkoalexandra Congrats on the launch! Can Claude save things back to Paste, or is it only able to search clipboard history?

Hey there,

Paste helps people save, search, and organize everything they copy across Mac and iOS. Today, we’re launching Paste MCP, a new way to bring your clipboard history into AI tools like Claude, Codex, and Cursor.

AI tools work best when they understand what you’re working on. But that context is often spread across links, notes, screenshots, snippets, files, and random ideas you copy during the day. Paste MCP brings that context into your AI workflow through a built in local MCP server on your Mac.

You can ask your AI tool to find something you copied earlier, use saved context in a draft, or create a pinboard without leaving your chat.

A few examples:

• Check Paste for the meeting notes I copied earlier and turn them into a team update.

• Find the links I saved about onboarding and summarize the key points.

• Create a pinboard for this project and add the relevant items.

Paste MCP runs locally on your Mac. You choose which AI tools can connect, and you can remove access anytime.

Big thanks to @chrismessina for hunting Paste again and supporting us on this launch.

We’d love to hear what you try first, what you find useful, and what you think is missing.

Thanks for checking it out!

Sounds cool! Definetely gonna try it, congrats on launch!

Сlipboard history as AI context, that's a brilliant reframe. been a paste user forever and never thought about it this way. the amount of stuff i copy daily and never look at again is wild in retrospect.

The MCP integration is a smart move — clipboard history becomes searchable long-term memory for Claude/Codex instead of losing everything when the session ends. One thing I wonder about is how you handle sensitive tokens or passwords that accidentally get copied. Is there a way to exclude certain apps or patterns from being saved?

Been using Paste for a long time, great idea!

Congrats! This is one of those tiny workflow pieces that probably matters more than it looks.

Most of us using Claude/Codex end up re-pasting the same repo notes, commands, and half-written instructions all day. The interesting bit is whether Paste becomes a shared memory for the work or another drawer full of snippets.

How are you thinking about stale context — stuff that was useful yesterday but wrong today?

I copy so much stuff throughout the day and lose track of half of it. clipboard history as searchable context for AI tools is one of those things that sounds obvious but nobody else has done it

Congratulations! Does it work with windsurfing? P.S. I love your project, by the way, and I'm an old user.

Love the idea, I think that have never seen product like this here yet.

It's a very sad day for the freedom of expression enjoyed by cats everywhere--but a great day for us cat owners! What a great tool! I just wish I owned a Mac.

The live demo fighting back against my own chaotic key mashing is genuinely hilarious. Every remote worker with a cat has lived this exact nightmare, and watching an app distinguish between "paw roll" and "human typo" in real time is just perfect. :p

Thank you for building something that protects both our Slack channels and our cats' reputations. This is the kind of joyful, useful tool that makes the internet better. 🐾

As a person with a missing key on my laptop from a cat attack - this is a great idea! Will there be a Windows version available at some point?

As a cat owner (check profile photo for proof) I approve this product! Such a great idea. My cats absolutely love plopping on my keyboard while I am working.

I don't have a cat, but this is so cute! Congrats on the launch 🐈‍⬛

free, open source, 100% on-device, solves a problem that has caused at least three incidents on my machine in the last month. this is the correct scope for a menu bar app

LOL I was one of the early user @ Toggl, now I'll support pawpause as well :D

the live demo where you try to type like a cat and watch it fight back is a great touch. also built the whole thing in a day with claude which is kind of the best ad for ai coding I've seen

Open source mac utility for the exact problem i never realized had a name. Bookmarking for the day i get a cat 😄

loved the name! waiting for this to be available on linux! :)

Finally, a cybersecurity product for the most dangerous insider threat: cats 😄 Congrats on the launch!

I can bet that LinkedIn would turn into a place full of shits :D

This is funny, but the underlying point is actually fair. The word “AI” gets added to so many things now that a small reminder to focus on the actual outcome feels useful.

Hahaha 😂 Good one, Wells. You genuinely made me laugh. Literally every website would be riddled with 💩 now.

Amazing. Gave me a good laugh reading some AI news. Reminds be of "Cloud to Butt" in a good way. :-)

the unintended consequence i'm thinking about is using this during a job interview or a client call where you're sharing your screen and forget it's on. has anyone shipped a 'disable for 30 minutes' button or is the chaos of forgetting it's installed kind of the point

The best thing that can happen for the current state we are in, thanks for making it. Soon to see AI language on the meat and chips in the stores...

@Ryan here from the TwelveLabs Product Team. 

The core idea behind Rodeo was that as it becomes easier to produce videos with the AI, the craving for real video will only increase and the real bottleneck will shift to finding the right clip in your footage. Editors re-watch the same footage on repeat. Marketing teams wait days for clip pulls. Agencies have no shared library. One natural language query solves all three.

A few things I’m especially proud of in this launch:

• Built from ground up on video-native multimodal ai that sees, hears, and understands your footage, not just the transcript

• Prompt-based editing allowing you to refine the cuts you make just by describing what you want

• Seamless handoff into Premiere, Resolve, or Final Cut with various NLE compatible export formats

• A generous free tier allowing you try it with your actual footage today, no commitment needed

We’re actively building and want to hear from creators on what to prioritize next. What’s your biggest video storytelling headache?

First launched Marengo 3.0 last December, then Pegasus 1.5 last April, award-winning @TwelveLabs keeps the momentum going with this new launch!

S/O ?makers 👏👏

The shared library problem for agencies is the most underrated part of this. Everyone has footage scattered across drives and Dropbox folders that nobody else can find or search. Congrats on the launch!

most video ai tools just work off the transcript and miss everything visual. being able to say 'find the shot where someone smiles at the camera' instead of scrubbing through hours of footage would save me so much time on marketing videos

Wow, wow, wow. Branda is beyond amazing. I opened an account expecting a claude design wrapper. I cannot believe how well designed, built and customised everything is. It generated my perfect branding system that I will be using for my own startup launching soon. beautiful work Eddie and David, you guys should be super proud of this. So stoked to see how Branda evolves!

Congrats on launching! The "expand on what you have" path is the one I'd want to test as a solo founder with an existing brand identity I'm happy with. Two questions on that mode: when I import existing assets, does the tool preserve them exactly when generating variants and mockups, or does it regenerate based on its interpretation? And is there a way to "lock" the imported core so that drift doesn't accumulate across many generations? The risk I'd want to mitigate is paying credits to slowly homogenize away from a brand I've already worked hard on. Will definitely take this for a spin :)

We're stoked to finally share this with the world! We hope you enjoy it as much as we have! ✨

Thanks for all the support everyone has given us over the last few weeks, we'd love to hear your feedback and do share on X some of the results you guy get (tag @brandacooks)

Love this! Super helpful for small teams. I'm the only marketing person at the current startup and I think this would be really helpful!

Brand systems are funny because the work is rarely one logo.

Its the messy stuff after: names, colors, screenshots, launch notes, weird little rules you forget until the next asset. If Branda keeps those decisions in one place without turning into a giant brand bible, thats helpful.

I’d be curious what you treat as the source of truth when a brand changes mid-project.

Very cool @lobanovskiy @kovadave ! Upvoted :)

Quick question: How many variants of logo/assets are created in one run? I see that each run is taking some 40 credits? What if the generated one was not satisfactory, can we redo for same credits? Thanks!

Hey Product Hunt 👋, I’m Phil, founder and Chief Chocolate Officer of choclift 🍫

It all started with a simple question: After 40 years, is there room for an additional way to interact with our Macs besides the keyboard and mouse/trackpad? Something that makes multitasking quicker, more direct and natural, and turns common keyboard shortcuts and endless Cmd+Tab cycling into something... sweeter?

Turns out there is. So we built it and better yet, we found a way to make it free.

choclift is the delicious chocolate bar on your iPhone that you can customize with your favorite apps, Apple Shortcuts, and websites, all launchable with a single tap. Instead of remembering keyboard shortcuts or navigating through layers of menus and bookmarks, choclift gives you a more direct way to interact with your Mac. Drag and drop feels instant, your favorite apps are always within reach, and new interactions like hiding apps become possible through simple swipe gestures. With a strong focus on interface design and user experience, choclift is unlike anything else out there.

And we’re only getting started. We're constantly exploring what else can sweeten a workflow. That’s why we added more swipe gestures for common actions like maximizing windows, copying and pasting, and even a real-time timeline of your recently opened apps. We call it “App Time Travel", an even quicker way to switch between apps.

If you’re skeptical about how this could fit into your workflow, give it a try. Maybe, without even realising it, you’ve been craving more sweetness while working with Mac all along: www.choclift.com

The App Time Travel framing alone made me smile. After decades of Cmd+Tab muscle memory, the idea of swiping through a visual timeline instead of cycling blindly feels like something my brain has been waiting for without knowing it.

Also, “Chief Chocolate Officer” might be the most honest job title on Product Hunt today. Congrats on shipping something that makes the Mac feel genuinely fun again; we need more tools that prioritize delight alongside utility.

very cool!

The branding here got me curious and then when I tried it out, it was a ... sweet treat... A really nice way to interact with the mac and make shortcuts and gestures easily accessible.

I really like pairing it with dictation and I def recommend for folks who have some sort of MagSafe stand next to their computer. It prevents the phone from being a distraction and turns it into an intentful controller for you Mac.

Turning the iPhone into a chocolate bar launcher for Mac apps and Shortcuts is a fun take on something I actually fight with daily (endless Cmd+Tab cycling). The branding made me smile too. Congrats on the launch Phil, followed you on X to keep up with what you're building.

the use case that would make me reach for this over keyboard shortcuts is something that keyboard shortcuts handle badly, like switching between more apps than you can memorize shortcuts for or triggering shortcuts that require multiple keystrokes to remember. are your early users mostly heavy app switchers or heavy shortcuts users because those might actually be different products with different retention patterns

Hey Product Hunt! Vlad here, indie dev behind GlowPulse 🩷

A year ago I noticed I was strapping an Apple Watch on every morning mostly to see my heart rate during work – to know when stress was climbing before I burned out. Then I’d take it off at night, lose it, forget to charge it. The whole wearable dance for one number.

Meanwhile my Mac has a camera staring at me all day.

Turns out there’s a whole research field called rPPG (remote photoplethysmography) that extracts heart rate from microscopic color changes in your skin caused by blood flow. Apple Watch does this with green LEDs on your wrist. A webcam can do it from across the screen.

So I built it. Started with a Python prototype using the POS algorithm (Wang et al. 2017), then rewrote it in Swift with Vision face landmarks and Accelerate FFT. Added an optional DeepPhys CoreML backend for ML-driven extraction. Benchmarked it against my Apple Watch – within ±3 BPM while stationary.

What ended up shipping is more than I expected:

• Live BPM in the menu bar with a sparkline

• Focus sessions (Pomodoro with HR charts – see stress climb during deep work, not after)

• Breathing sessions with live HRV at 0.1 Hz coherent breathing

• 30-second stress checks against your personal baseline

Privacy was non-negotiable from day one: every camera frame is processed in memory and discarded. No cloud, no account, no telemetry. The entitlement list is literally just “camera.”

$2.99 once, lifetime updates. 7-day free trial on the direct download.

Wildly open to feedback – it’s v1.0 and there’s still a lot to improve. AMA!

This is a really novel idea, did you cross-validate this with another type of HR sensor, such as your Apple watch or an over-chest wearable?

Wow, had no idea this was possible. Awesome, why doesn't everyone just use this?

How accurate is it when lighting changes or you’re moving a bit?

That is so cool honestly!

Do you have another way to measure your heart rate? Did you see any difference between the camera and your sensor?

Neat piece of software but does not track very well and show high drift compared to my Garmin Fenix 7 ( M4 Pro 16" mbp, Avaya HC020 camera) - will revisit in a few months. Would be nice to add a posture tracker, breathing tracker)

Hey Product Hunt — Chris from Knock here.

We built Knock agent for Slack because customer messaging work already starts in Slack: someone asks for a launch email, a lifecycle update, an audience change, or a campaign tweak, and then the actual work has to move somewhere else.
Now you can tag @Knock in Slack, describe what you need, and let the agent work in the background to build, edit, and manage your customer engagement resources.

It can help build campaigns, draft messages, update workflows, and answer questions about your messaging setup. Once it’s done, it provides a direct link to review any modified resources in Knock.

The important part: the agent gathers the context from your account—your workflows, audiences, layouts, styles, channels, and brand system—so the output is always on-brand.

We see this as a step toward agent-led customer engagement: less tool-switching, fewer repetitive edits, and faster movement from idea to shipped messaging.

Would love feedback from anyone building lifecycle, product, growth, or customer communication systems.

Knock on,

Chris Bell, CTO + Co-founder @ Knock

The team cooked on this one! We're excited for you all to try the Knock agent in Slack and to bring lifecycle messaging infrastructure to all the surfaces where you want to use it. Try it and let us know what you think!

This launch is a special one. For a couple weeks now, I've been messaging Knock from Slack to build our May newsletter, update our onboarding campaign, draft in-app messages, and even create the broadcast for this launch.

It's as simple as a tag and a prompt...and I get a link back to review and publish.

If you’re excited about what we’re building, we’d appreciate your upvote. Try Knock today to see it in action 🚀

building campaigns by just tagging an agent in slack instead of switching to another dashboard is how more tools should work. the context is already in the conversation so why leave it

Love this idea. If conversations start in Slack anyway, it makes sense that campaign creation should happen there too.

Huge congratulations to the Knock team! Reducing context switching is one of those improvements that sounds small but has a huge impact on productivity.

This feels like a great example of AI removing busywork rather than creating more of it. Nice work by the team!

Love the privacy-first approach. No accounts, no servers, no subscriptions — that's becoming surprisingly rare these days. Congrats on the launch!

I agree that this is a real problem for SaaS teams. Users can sign up with interest but if the first few steps feel confusing, they may leave before seeing why the product is useful. I like that Kompassify is not only abt product tours but also abt seeing where people drop off. How does the AI copilot decide what should be improved first when there are a few weak points in the onboarding flow?

Congrats on the launch! I saw you mention that teams can ship onboarding much faster with Kompassify than with the usual tools. I’m curious how that plays out in practice is it mostly the no‑code editor, or did you cut out some of the heavy configuration other tools force you into?

Love it, people spend so much on "acquiring" customers and not enough on actually keeping them coming back those first few days/weeks.

Context quality seems to be becoming a bigger bottleneck than model quality for a lot of agent workflows.

Are you seeing teams use Moxie primarily as documentation, or more as a structured context layer for coding agents?

Docs that travel into the coding assistant is the part that feels right here.

A README sitting in GitHub is helpful, but the moment I'm in a Claude/Codex loop the question is whether the current rule is actually in the model's face. I’d be curious how you handle closed decisions — the architecture rule that should not be re-litigated every new chat.

I had thought that Product Hunt would add the promo automatically since I entered it (maybe I just missed it) but you can also use PRODUCTHUNT on checkout for 50% off a month!

Tested it on my two apps and immediately found something App Store Connect never made obvious: the gap between impressions and page views told me one app is discovered almost entirely through search while the other lives off external traffic. That changes how I think about each one. The pre-defined stats get you value in minutes, then you can dig into the raw reports. Nice work Florian.

As someone who's squinted at App Store Connect knowing the data I want exists somewhere but isn't surfaced, unlocking all 100+ report types is genuinely handy. Love that it pulls on-device data too without a third party. Nice work Florian, followed you on X to keep up with your apps.

Finally, a tool for the thousands of screenshots I swear I’ll organize “later.” 😅 The local OCR + privacy-first architecture makes this stand out. Looking forward to trying it out!

rust/tauri + native Vision — nice combo. built similar local-first on mac, the OCR pass got memory-heavy past a few thousand shots. you indexing incrementally on new ones or batching the lot?

I like the local-first approach. Is Mirowl currently macOS-only, or are there plans for a Windows version in the future?

Built a K1 document OCR pipeline professionally and spent the last week generating hundreds of AI video storyboard screenshots across GPT and Kling. The screenshot search problem is real — I was manually scrolling through folders looking for specific reference images. Curious how Mirowl handles images with minimal text — like a cinematic still or a product photo with just a logo. Does the AI understand visual content beyond just OCR, or is it purely text extraction from images?

Nice one @safiullah_mohamed , upvoted :)

When you say local OCR powered AI, how much extra infra would it need to run?

Nice that everything happens on device, as i state with thoth, our devices have more computing power than the Apollo guidance computer so why offload computing to the cloud !

Quick question: does Mirowl rename the files or propose titles based on the content, or is it search-only? Auto-naming "Screen Shot 2026..." into something findable would be huge for me. To keep that local it could be done with apple intelligence

Local OCR is a smart call, especially for screenshots that often have personal stuff in them. Curious which engine you went with under the hood, Apple Vision or something custom? I work with on-device OCR too and the accuracy on dense text was the hardest part to get right.

Hey Product Hunt! 👋

  I'm a photographer, so this is a problem I lived with for years.

  THE PROBLEM

  After every shoot, my photos were already sitting in Google Drive. But when it was time to deliver

  them to the client, I had two bad options:

  1. Share a raw Google Drive or WeTransfer link. It works, but the client opens a messy file list,

  often gets confused, and the experience feels nothing like the quality of the actual work.

  2. Re-upload every single photo to a dedicated gallery platform. Clean result, but it means

  duplicating gigabytes of files and wasting time after every single job.

  Neither felt right. My work looked great, but the delivery didn't.

  THE SOLUTION

  So I built Galleroo. The idea is simple: your photos already live in Google Drive, so why move them at

  all?

  How it works:

  1. Create a folder in your own Google Drive and drop in the client's photos and videos (exactly like

  you already do).

  2. Connect that folder to Galleroo and give the gallery a name and a password.

  3. Share one link with your client.

  In under 60 seconds, that same Drive folder becomes a clean, full-screen, professional gallery.

  WHAT YOUR CLIENT GETS

  • A beautiful, responsive gallery that works perfectly on phone and desktop

  • No account or signup needed. They just open the link.

  • They can mark favorite photos and filter to them instantly

  • They can download in full resolution, with no compression

  • Password protected, so only the right people get in

  WHAT MAKES IT DIFFERENT

  • No re-uploading. Your photos never leave your own Drive. Galleroo simply presents them.

  • No subscription. One-time annual payment, no surprises.

  • Other tools force you to upload everything again to their servers. Galleroo doesn't.

  I built this because I believe a photographer's work deserves to be presented as well as it was shot.

  I'd genuinely love your honest feedback! ❤️

This solves a real small-client handoff problem.

Google Drive works until the client starts asking which folder, which version, which images are approved. The gallery layer is probably less about looking polished and more about reducing those little back-and-forth messages.

How are you handling approvals or selected images inside the Drive flow?

I like this idea! Would this also work with other file types, like .PDF or .docx? Sometimes I want to share a folder with colleagues from my Drive but I don't necessarily want to deal with permissions and access rights - this could be a good solution for that!

Real use case for this - built an AI film for a product called Dreams of Yesterday that deliberately has no subtitles during a key flashback sequence. The decision was intentional — wanted the disorientation. But for accessibility and multilingual families watching together, real-time captions without modifying the video file would be exactly right. Curious how Overline handles videos with mixed dialogue and silence — does caption accuracy drop when audio is sparse or ambient rather than continuous speech?