News for hackers. Stuff that matters.

fliellerjulian

Show HN: AthleteData – AI coach for endurance athletes that messages you first

https://www.athletedata.health

Im a triathlete and the data for my training lives in 6 apps: Garmin, Strava, WHOOP, Intervals.icu, Wahoo, Withings, Apple Health, sometimes Hevy.Every morning Id eyeball a few of them and make a call on whether to do the planned session. For the past month I have been building a thing that does this for me, and got it to the point where I use it myself every day.It OAuths into whatever...

( Read More... | no comments | 5 points | on HN )

Show HN: AI agent forgets user preferences every session. This fixes it

https://www.pref0.com/

I build AI agents for work and kept hitting the same issue: a user corrects the agent, the session ends, and the correction is gone. Next session, same correction. I tracked it across our users and the average preference gets re-corrected 4+ times before people just give up. Existing solutions don't really solve this. Memory layers store raw conversation logs. RAG retrieves documents. Neither...

( Read More... | no comments | 6 points | on HN )
Re:Mem0 stores memories, but doesn't learn user patterns on Thursday February 05, 2026 @03:55AM (#46897401)
100% feel the same way. I´ll look into building an API for that so more people/companies can use it.
Re:Mem0 stores memories, but doesn't learn user patterns on Thursday February 05, 2026 @03:54AM (#46897397)
oh my bad, thanks for flagging. I dont have a link to anything thats why I did do Show HN
Re:Mem0 stores memories, but doesn't learn user patterns on Wednesday February 04, 2026 @04:08PM (#46891797)
we did the same thing, but we noticed most users did not maintain their own system memory themself properly. so we had to build a solution that auto-manages memory and preferances.

Mem0 stores memories, but doesn't learn user patterns

We're a YC W23 company building AI agents for engineering labs - our customers run similar analyses repeatedly, and the agent treated every session like a blank slate.We looked at Mem0, Letta/MemGPT, and similar memory solutions. They all solve a different problem: storing facts from conversations — "user prefers Python," "user is vegetarian." That's key-value memory with semantic search....

( Read More... | 7 comments | 3 points | on HN )

Show HN: PR comments for Claude Code diffs (open-source)

https://github.com/fliellerjulian/claude-code-feedback

When Claude Code makes changes, giving targeted feedback is annoying. You have to switch back to chat and explain "actually, change line 47-65...."I built a VS Code extension that adds inline comment boxes directly on git diffs. Click a line or select multiple, type what you want changed, hit send -> goes straight to Claude Code's terminal with the exact diff context.Makes the feedback loop...

( Read More... | no comments | 3 points | on HN )
Re: on Wednesday April 16, 2025 @10:09PM (#43712391)
Last month, we launched Openspot, a more human alternative to LinkedIn, designed to help job seekers stand out through short video/audio intros and storytelling prompts. Think: YC-style application, but for the real job market.

We hit #1 on Hacker News and Product Hunt and got our first 200 users — early traction has been encouraging.

But here’s the challenge we’re now facing: While some power users get it immediately, they upload video bios, fill out their profiles, and have already landed interviews — the majority just keep the pre-generated profile and bounce.

We built Openspot around the idea that resumes are becoming indistinguishable. With AI, every candidate “spearheaded cross-functional initiatives” and “drove impact across orgs.” It's harder than ever to tell who’s actually behind the text.

Our solution is to make hiring feel more human again, letting people show who they are, not just what they've done.

But now we're wondering: How do we help more users understand that this is the point?

Should we focus on educating them through better onboarding, incentives, and nudges?

Or should we double down on finding more people who already “get it” without much explanation?

Curious if anyone has gone through something similar — building a product that requires a behavior shift. How did you cross the chasm between early believers and the rest?

Would love to hear your thoughts.

fixed ^
idk why it has been flagged..anyways, the mobile display is now fixed ;)
just like linkedin, we will be charging recruiters to access the platform and search for candidates
That’s a fair comparison, but we’re aiming for something much deeper than just a pretty landing page.

Openspot is about curated professional identity. It lets people showcase not just what they do, but who they are — through video, audio, projects, and real storytelling. It’s built to help people stand out in a world where resumes are starting to look the same (especially with AI).

We’re also bringing in discovery tools for recruiters, so it’s not just a personal page — it’s a real alternative to the resume-black-hole problem.

LinkedIn became Facebook with resumes..you basically need a "personal brand" to get hired
Yeah Totally get that! for some people, sending a solid CV and letting the work speak for itself should be enough.

Openspot is more for folks who feel like their resume doesn’t tell the whole story. Maybe they’re career-switching, self-taught, or just want to show a bit more of who they are, not just what they’ve done.

That said, we’re working to make the process super lightweight — upload your resume, we generate a profile in under a minute. So it’s more of a bonus layer if you want to stand out, not a whole extra job search on top of everything else.

we have a "projects" section where you can upload, pdfs, images, links, and videos and also ofc write something about each project :)
looking into it, thanks!!
haha thanks!!! :)
totally agree on the last part

What’s wrong with LinkedIn (in our view): - It’s become more like Facebook than a professional network: lots of noise, humblebrags, and algorithm-chasing. - Everyone looks the same on paper — resumes blur together, and it’s hard to stand out. - You're incentivized to perform rather than be real — especially if you're early in your career or outside of the “right” networks.

How we're different: - No feed. No likes. No fake engagement. - You show who you are with video, audio, and proof of work — like a mini personal pitch. - Profiles are curated and designed to be browsed, not scrolled endlessly.

Re: network effects — We’re starting small and high-signal: curated profiles + inbound interest from early recruiters. Our goal isn’t to replace LinkedIn’s entire graph, it’s to become the place you go when you want to stand out and be discovered for real.

Since our initial launch, we have also launched in India, but we are currently working on all legal requirements to launch in more countries very soon!
The Frontend is build with React, while the backend is written in Python (Flask). For the Database, we use MongoDB and everything is hosted on AWS.

To build fast, we used cursor and some other third party SaaS like streamIO and Supabase.

Two weeks ago I posted "LinkedIn sucks, so I built a better one" here on HN — it unexpectedly hit #1, and the feedback was incredible.

Since then: - we hit #1 on Product Hunt the next day - got featured in their daily + weekly newsletter - 1,500+ sign ups - 1,000+ waitlist sign ups from non US countries - 32,000+ website visitors - 35 recruiter signed up

The response gave us so much conviction that we decided to go all-in: we left our Big Tech jobs to work on Openspot full time.

It still feels surreal.

What we are building: In a world where AI can write the perfect resume for anyone, what actually makes you stand out is you — your personality, your voice, your story. That’s what OpenSpot helps you showcase — through video, audio, projects, and more.

If you're curious what we've learned so far — from launch tactics to handling the wave of attention to quitting Big Tech jobs to build this full-time — happy to share everything.

Re:Show HN: LinkedIn sucks, so I built a better one on Tuesday March 25, 2025 @09:22AM (#43470908)
its based on screen width, thats why!
Re:Show HN: LinkedIn sucks, so I built a better one on Tuesday March 25, 2025 @09:21AM (#43470893)
Couldn’t agree more — this is definitely a cold start + network effects problem at its core. We’re under no illusion that we can out-network LinkedIn by being shinier or faster.

You nailed it: their true moat isn’t tech, it’s scale and inertia. Everyone’s on it, even if they hate it. That’s what makes it both frustrating and interesting to challenge — not because it's easy, but because it’s overdue.

We’re not trying to beat LinkedIn at their own game (at least not yet) — we’re focused on solving one painful wedge: helping people actually stand out and get discovered without shouting into a noisy feed or spraying cold applications into the void.

If we can win on that one slice of user need (especially among early-career folks, designers, PMs, indie hackers, etc.), we believe there’s room to build up a meaningful alternative, piece by piece. Small, high-signal networks that grow from the edge rather than trying to eat the core from day one.

Appreciate the thoughtful challenge — these are exactly the conversations that keep us grounded and ambitious at the same time.

Re:Show HN: LinkedIn sucks, so I built a better one on Tuesday March 25, 2025 @09:17AM (#43470843)
Wow — thank you for this. You totally get the pain we’re trying to solve, and I really appreciate the encouragement

Funny enough, “The Anti-LinkedIn” has come up more than once during this launch - and I agree, the copy kind of writes itself when people feel the same pain.

Also +1 on Huntr — great tool! We’re focused more on visibility and standing out before the apply button, but I think they complement each other well.

You nailed something really important: the real opportunity is to create networks with actual signal — thoughtful discussions, deep expertise, and real connection. Not the dopamine feed. We’re starting super lightweight, but that’s absolutely part of the long-term vision.

And yeah… LinkedIn Learning + random recruiter spam + engagement farming = what pushed us to build this in the first place. Thanks again for taking the time — would love to stay in touch as we build.

Re:Show HN: LinkedIn sucks, so I built a better one on Tuesday March 25, 2025 @09:10AM (#43470762)
Haha fair — it’s kind of in between a resume and a portfolio.

OpenSpot isn’t meant to replace formal resumes entirely — it’s designed to help people stand out before that stage. It’s public, searchable, and curated — so if a recruiter finds someone interesting, they can reach out directly or request a formal resume.

Photos are optional, but yeah — they can help distinguish one Jane Doe from another, especially in a directory-style setting. Same goes for a short intro, project link, or anything else that brings some personality into the mix.

Re:Show HN: LinkedIn sucks, so I built a better one on Tuesday March 25, 2025 @07:51AM (#43470125)
Totally hear you — and I really appreciate how thoughtfully you laid this out.

Photos are not required on OpenSpot - We included them because some people want to humanize their profile — but we’re also very aware of the bias/discrimination concerns, especially in the U.S.

The intent isn’t to mimic LinkedIn or create a polished image-centric profile like a dating app — it’s to offer optional ways for people to show who they are, beyond just a list of bullet points. For some, that’s a photo. For others, it’s a project demo or a short intro about their work.

That said, we’re actively listening and learning. If we find that photos do more harm than good — especially around bias or UX — we’re open to evolving that part of the product.

Thanks again for the thoughtful take — this kind of feedback makes the product better.

Re:Show HN: LinkedIn sucks, so I built a better one on Tuesday March 25, 2025 @07:49AM (#43470116)
Totally valid — if LinkedIn is working for you, that’s awesome.

The reason OpenSpot might "suck less" for some is because it focuses only on one thing: helping people stand out. No feed, no fluff, no algorithm — just a clean, curated profile where you can show off your skills, projects, and intent in a way that cuts through the noise.

And to be clear: you don’t have to record a video. That’s optional. Some people link to projects, write a short blurb, or just highlight their strengths in a more human way than a traditional CV.

It’s not a LinkedIn replacement — it’s an alternative for those who feel like they’re getting buried in the system and want more visibility without the social media layer.

Appreciate the honest question!

Re:Show HN: LinkedIn sucks, so I built a better one on Tuesday March 25, 2025 @07:46AM (#43470106)
Totally fair point — and one we’ve thought a lot about.

The goal with OpenSpot isn’t to force everyone to do more unpaid work or replace resumes entirely — it’s to give people who want to stand out an easy, high-signal way to do so, especially if they’re not getting traction through traditional means.

For recruiters, the platform is curated and searchable, so they’re not sifting through hundreds of TikTok-style intros — they’re seeing real signals of quality, with depth if they want it.

We're not trying to increase the noise — we're trying to cut through it.

Re:Show HN: LinkedIn sucks, so I built a better one on Sunday March 23, 2025 @06:24PM (#43456383)
thats awesome, looking forward to your feedback! feel free to join our discord so we can keep in touch

Older »