Comments URL: https://news.ycombinator.com/item?id=45366538

Points: 2

# Comments: 0

I built Kortyx after a failed attempt to find product-market fit with an AI product-analytics startup.

When it was time to pivot i was looking for other ideas and figured that i constantly struggled with getting users to provide "ideal" context to my agents. it resulted in weak prompts, degraded agent performance, etc etc.

So i thought what is instead of asking every user to provide background, Kortyx captured the context directly and then served it at the right time.

Kortyx is a desktop app (Windows + macOS) that quietly sees what you see and builds a private memory layer from your digital life. You can:

Ask Memory – recall anything you’ve read, watched, or discussed with an exact snippet and timestamp.

Memory Boost – superpower your prompts to other AI agents with detailed context about you.

I’m launching today and would love feedback, especially from Mac users and anyone exploring agent ecosystems.

Comments URL: https://news.ycombinator.com/item?id=45290846

Points: 1

# Comments: 0

Hey, I'm Dima, a former PM who led product redesign and had a constant analytics hangover, and now a solo SF founder who only has time either to ship a new product or recruit users. Analytics is always lagging behind and piling up into an annoying debt.

I was also reflecting on product management as a job in general and was thinking about hard skills, and came to a conclusion that analytics is the only hard skill of a PM, and analytics is essentially the coding of the PM.

Given the AI and MCP boom, I figured it's possible to automate even that. So I built UserWatch: an AI product analyst that connects to PostHog and sets everything up for you. (long warm up before pitch ik, sry) You just tell it what you want in plain language, “show me the onboarding funnel” or “run an A/B test on the pricing page” and it does the rest.

No tagging, no waiting for your analytics team, no manual setup. It just works(well, in beta, lol). If you’re a founder or manager who needs data to decide where to stir the product dev - check it out! Happy to set you up for free if you want to try it. Open to any questions :)

Comments URL: https://news.ycombinator.com/item?id=44249435

Points: 2

# Comments: 0

It's just cursor's system prompt problem. they just need time to "tame" the model after release.

for now, I just make sure that in every chat thread i have "DO NOT WRITE ANY DOCUMENTATION OR TEST OR ANYTHING THAT WASN'T EXPLICITLY ASKED. STAY LEAN"

But i sort of reached the point where I don't mind claude going off the rails a bit. Like restricting it with .md and constantly updating those guardrails sounds like more of a burden than help.

its just the prompt problem. try reading the chat and every time you see it doing excessive shit stop the chat and slap it on wrist saying "never create .env i already have it, you just don't have access, etc".

also, sounds obvious, but don't forget to create new conversations often. the "ignore a files it's just created" sounds like context window overload. 200k window for a new model sounds like a crime from anthropic

Something that's been working for me lately is to choose the topics where you have something to say. It's a bit broader than the things you like and allows you to just react to an inner spike to respond. Helps train the muscle for writing

Why working on this So I'm a former PM, and product analytics seemed like only semi-"hard", transferable skill one can have. But the problem is that for every company, you still have to find the key metrics anew and the process of finding the right way to track everything takes multiple attempts and is slowed down by having to wait for Analyst to prepare date for you and for developers to set up event tracking and ab test displaying. In addition to all that, when you release a new feature and it doesn't work - you have to sit and stare at goddamn session recordings 90% of which are mind-numbing and not insightful.

That entire process is essentially the "coding" skill of PM. And that's what I'm automating with UserWatch and its Watchy AI.

How it works Right now, UserWatch integrates with Posthog and can help you with:

Dashboards on demand – just prompt info you want to see. "activation funnel where users visit page X, then click button with text Y, and then trigger the event Z". Watch AI will confirm whether it identified every action correctly and send you a link for the dashboard.

Set up A/B tests – “roll new onboarding to 30 % of traffic, success metric is sign‑up” → Watchy will create a flag feature + experiment.

Replay analysis – Describe the funnel step where you notice people drop off and userwatch will conduct a deep research to provide you the reasons why it happens and the Jira fix tickets.

Stack AWS, rrweb for replay capture, PostHog for analytics backend, gpt-4.1, agents sdk, posthog custom mcp.

Try it free 14-day trial with 15k sessions analyzed. Looking for brutal feedback on:

What you failed to setup with prompts, but really want to

Anything that would stop you from replacing GA4(god forbid), PostHog, etc.

Happy to answer technical questions. Non-technical too. Some personal too. Love ya!

Comments URL: https://news.ycombinator.com/item?id=44159317

Points: 1

# Comments: 0

I used to be the PM responsible for onboarding and activation. After hundreds of hours binge‑watching session replays and running user interviews, I realised we were still answering “Why did this user ghost after the trial?” on vibes.

So I built *UserWatch*, an AI system that:

1. Captures plain‑vanilla session replays (drop one script tag). 2. Re‑watches every session and tags each exit as one of *7 root‑cause types* (UX confusion, value gap, tech error, etc.). 3. For each recurring pattern it: • bundles the best 30‑second clips to prove the hypothesis • suggests a concrete UI / flow tweak to move those users to the next funnel step.

*Why it’s useful*

* No more hunting for “the good replays” or guessing what happened. * You get a ranked list of drop‑offs, evidence, and ready‑to‑ship fixes for the next sprint. * Caveat: analysis runs in a 24‑hour batches—tokens aren’t free and neither is my OAI bill.

*Status*

Private beta. I’m hunting for a handful of SaaS teams willing to break it and tell me what’s wrong. If that’s you, I'm excited to chat.

Happy to take any questions—brutal honesty welcome :)

Thanks! Dima

Comments URL: https://news.ycombinator.com/item?id=43895649

Points: 1

# Comments: 0