frontpage.
newsnewestaskshowjobs

Open Source @Github

fp.

Multiple agents, two Macs, and several skills

https://www.swiftjectivec.com/the-skills-conundrum/
1•jordanmorgan10•36s ago•0 comments

Before Rightmove, there was the Cosmorama

https://www.ianvisits.co.uk/articles/before-rightmove-there-was-the-cosmorama-londons-forgotten-p...
1•brod_ie•1m ago•0 comments

Show HN: Steno – static site generator with sandboxed plugins and atomic builds

https://steno.gxbs.dev
1•immjs•2m ago•0 comments

Simple Sudoku Solvers

https://blog.veitheller.de/sss/
1•gregsadetsky•2m ago•0 comments

University of Michigan Drops First-Semester Grades To'Curb Mental Health Crisis'

https://www.wsj.com/us-news/education/university-of-michigan-grades-mental-health-1a5701d4
2•cwwc•2m ago•0 comments

Show HN: VitCam – Self-hosted, open-source NVR with on-device AI detection

https://github.com/scwsoft/vitcam
1•scwoods•3m ago•0 comments

Bernie Sanders calls on Silicon Valley to 'pause AI development'

https://www.theguardian.com/technology/2026/aug/10/bernie-sanders-ai-development-pause-letter
2•reducesuffering•3m ago•0 comments

Show HN: Username.md – a signed, agent-readable identity page you own

https://username.md/
1•cyberge99•3m ago•0 comments

The Dartmouth Workshop–as planned and as it happened

https://www-formal.stanford.edu/jmc/slides/dartmouth/dartmouth/node1.html
1•Bluestein•3m ago•0 comments

Rebundling the database into a language

https://www.cambra.dev/blog/rebundling-the-database/
1•altairprime•3m ago•0 comments

Show HN: Kessa, a verifiable delegation and attestation system for AI agents

https://github.com/Gneiss-Group/Kessa
1•raphtheb•4m ago•0 comments

A New Type of Levee Is Taking Root in Northern California

https://grist.org/solutions/a-new-type-of-levee-is-taking-root-in-northern-california/
1•thatoneengineer•5m ago•0 comments

Zoomsday How a Security Found a Nation-State Vulnerability in Zoom in One Day

https://a.security/blog/asecurity-zoomsday
3•shscs911•6m ago•0 comments

Silent Data Corruption in PyTorch

https://github.com/pytorch/pytorch/pull/192575
1•carterschonwald•6m ago•1 comments

Electric is joining team Neon at Databricks

https://neon.com/blog/electric-joins-neon
2•bayandin•7m ago•0 comments

Show HN: My Anti AI Computer

https://punkx.org/jackdoe/home.html
1•jackdoe•7m ago•0 comments

Multi-tenant BYOK encryption in PostgreSQL with pgcrypto

https://xata.io/blog/multi-tenant-byok-encryption-in-postgresql-with-pgcrypto
1•tudorg•7m ago•0 comments

WAL Levels in Postgres and Effective WAL Level in PG19

https://www.buraksen.dev/articles/wal-levels-in-postgres-and-effective-wal-level-in-pg19
1•buraksen•7m ago•0 comments

Degrees of Freedom

https://exple.tive.org/blarg/2026/08/10/degrees-of-freedom/
1•CountHackulus•8m ago•0 comments

Can Agents Use a Computer Yet?

https://www.a16z.news/p/can-agents-use-a-computer-yet-weve
2•deepmem•9m ago•0 comments

Hexis: Git-backed skills, tools and context for AI agents

https://github.com/Bevel-Software/Hexis
3•juanviera23•10m ago•0 comments

DL591 LAS-ATL arrival met by federal agents following mid-air WiFi hacking

https://www.reddit.com/r/delta/comments/1vl52vr/dl591_lasatl_arrival_met_by_federal_agents/
2•_____k•11m ago•0 comments

Show HN: Pickle – token efficient AI agent browser with policy-gated actions

https://picklebrowser.com/
3•ismail_h•13m ago•0 comments

Mother of all VPNs v2 release: Same 16 protocols, rebuilt underneath

https://twitter.com/MotherofallVPNs/status/2087169933604008016
2•shayanbahal•15m ago•1 comments

We built our own status page with AI, replacing a $65k SaaS product

https://www.fivetran.com/blog/we-built-our-own-status-page-with-ai-replacing-a-65k-saas-product
2•georgewfraser•15m ago•0 comments

Jackie the bald eagle dead at 14

https://www.popsci.com/environment/jackie-bald-eahttps://www.popsci.com/environment/jackhttps://w...
2•cyndunlop•16m ago•1 comments

Nvidia Nemotron 3.5 Lightning

https://twitter.com/NVIDIAAI/status/2087162151995629926
3•tosh•16m ago•0 comments

What does International Vlogging Day mean in 2026?

https://sijobling.com/micro/what-does-international-vlogging-day-mean-in-2026/
2•mooreds•17m ago•0 comments

PDCA

https://en.wikipedia.org/wiki/PDCA
3•tosh•17m ago•0 comments

AI Is Accelerating Vulnerability Discovery

https://ai-updates.net/ai-accelerating-vulnerability-discovery-inventory/
2•ashurandi•18m ago•1 comments
Open in hackernews

Relevare: Self-driving AI enablement and modernization for non-technical teams

https://relevare.tech
1•npb12•53m ago

Comments

npb12•53m ago
It’s been said that “AI has a multiplying effect on existing technical skills” (https://news.ycombinator.com/item?id=48235526), but what about everyone else?

There’s a massive productivity gap growing between technical professionals and the rest of the workforce, and that gap is growing every day. Though we have access to the same tools, a chat interface doesn’t come with an engineer’s mind or an architect’s vision for solving problems. Relevare was built to close that gap and give every professional the chance to become 10x at their job.

How it works: Relevare runs on Mac and Windows and learns a user’s role, what they’re trying to achieve, and what their day-to-day work looks like. Solution architects and research agents analyze those workflows to find bottlenecks and opportunities that could make the user more productive and successful. That could mean suggesting better practices for a salesperson prospecting with Claude and LinkedIn, or designing an automation for a construction project manager that flags schedule risks and overdue RFIs. AI is not always the answer. Relevare might find a process that should be simplified, an existing feature that is not being used, or a tool that is better suited to the job. Other times, the right answer might be a traditional automation, a custom application, or an AI-assisted workflow. When a custom solution is the right answer, Relevare can send it to ChatGPT or Claude to be built with one click (More integrations are coming).

Our goal isn’t to turn users into “vibe workers” who hand their jobs to AI and lose all control. That’s why each new workflow that’s introduced comes with enablement and continuous checks on the latest and greatest ways to perform a task. The user remains in control of what changes.

Because understanding someone’s work requires a lot of context, privacy was a core part of how we built Relevare. Raw screen captures stay on the user’s device. Relevare is not designed to score employees or give managers a view into individual activity. When used across a team, company leaders see shared workflow patterns in aggregate, not which employee they came from.

Some of our early users had never automated anything beyond a credit card payment. We’ve watched them begin to improve processes, discover features in tools they already use, and selectively introduce automation and AI when applicable.

Although Relevare was built primarily for non-technical professionals, some of our strongest early validation has come from engineers who found that their own work could be improved too. The standard of work has changed faster than we’ve ever experienced. Relevare is meant to give everyone a chance to keep up, regardless of their technical background. We’d love to hear what HN thinks. Your can try it for free with your ChatGPT account.