Feed aggregator

Article URL: https://www.hackster.io/news/nino-ivanov-brings-dec-s-pdp-8-minicomputer-out-of-the-60s-and-onto-an-arduino-near-you-c7745e17e7fa

Comments URL: https://news.ycombinator.com/item?id=47230448

Points: 1

# Comments: 0

EU legislation (which affects UK and US companies in many cases) requires being able to truly reconstruct agentic events.

I've worked in a number of regulated industries off & on for years, and recently hit this gap.

We already had strong observability, but if someone asked me to prove exactly what happened for a specific AI decision X months ago (and demonstrate that the log trail had not been altered), I could not.

The EU AI Act has already entered force, and its Article 12 kicks-in in August this year, requiring automatic event recording and six-month retention for high-risk systems, which many legal commentators have suggested reads more like an append-only ledger requirement than standard application logging.

With this in mind, we built a small free, open-source TypeScript library for Node apps using the Vercel AI SDK that captures inference as an append-only log.

It wraps the model in middleware, automatically logs every inference call to structured JSONL in your own S3 bucket, chains entries with SHA-256 hashes for tamper detection, enforces a 180-day retention floor, and provides a CLI to reconstruct a decision and verify integrity. There is also a coverage command that flags likely gaps (in practice omissions are a bigger risk than edits).

The library is deliberately simple: TS, targeting Vercel AI SDK middleware, S3 or local fs, linear hash chaining. It also works with Mastra (agentic framework), and I am happy to expand its integrations via PRs.

Blog post with link to repo: https://systima.ai/blog/open-source-article-12-audit-logging

I'd value feedback, thoughts, and any critique.

Comments URL: https://news.ycombinator.com/item?id=47230438

Points: 1

# Comments: 0

Article URL: https://squadofagents.com/

Comments URL: https://news.ycombinator.com/item?id=47230407

Points: 1

# Comments: 0

Article URL: https://github.com/BankStrategy/SOEL

Comments URL: https://news.ycombinator.com/item?id=47230399

Points: 1

# Comments: 1

Article URL: https://www.writingruxandrabio.com/p/equality-as-a-consolation-prize

Comments URL: https://news.ycombinator.com/item?id=47230397

Points: 1

# Comments: 0

Article URL: https://www.euronews.com/my-europe/2026/03/03/how-close-is-the-eu-to-break-free-from-visa-and-mastercards-grip

Comments URL: https://news.ycombinator.com/item?id=47230391

Points: 1

# Comments: 0

Comments URL: https://news.ycombinator.com/item?id=47230384

Points: 2

# Comments: 0

Article URL: https://itnext.io/scaling-time-series-to-millions-of-cardinalities-greptimedbs-flat-format-7a9d67dd8000

Comments URL: https://news.ycombinator.com/item?id=47230380

Points: 1

# Comments: 0

Hey HN,

For a while now, I've been building a cognitive AI business operating system from the ground up natively as a solo developer. I never coded before 6 months ago. I was frustrated with wrappers and fragile single-agent setups, so I decided to approach this as a distributed systems problem instead. It's called *Sophify*.

Instead of trying to make one monolithic capability do everything, I built a 12-server, 28+ microservice swarm orchestrated by over 40 highly specialized cognitive AI agents.

*Here’s the architecture under the hood:* * *40+ Agent Architecture (Model Agnostic):* Decisions aren’t made by a single prompted agent. There is a multi-agent orchestration system that uses weighted voting and consensus across 40+ specialized cognitive agent profiles to synthesize decisions. Crucially, the system is LLM-agnostic—you can plug in whatever LLM you prefer (local or remote) to power the reasoning engines. * *The Technician Bot:* Given that this comprises 28+ interconnected microservices running in the cloud. I wrote a persistent Technician layer. If a node hangs or hits a port collision (hello `WinError 10048`), the Technician detects it, diagnoses the stack, kills zombie processes, reallocates networking, and brings the service back online with zero human intervention. * *1,600+ Industry Verticals:* Rather than building general-purpose workflows, the OS comes pre-loaded with state machines and hardcoded context for over 1,600 specific verticals running 1,284 autonomous workflows—handling everything from healthcare compliance to real estate out-of-the-box. * *Military-Grade Security & Built-in VPN:* The entire system runs inside an integrated, zero-trust VPN layer with cryptographic audit trails and multi-tenant isolation. Production-ready for DoD contracts, healthcare (HIPAA), and enterprise compliance. * *Deep Microsoft 365 Integration:* Full Microsoft Graph API implementation with native access to email, calendar, Teams, OneDrive, SharePoint, and Planner. The system orchestrates across AWS, Azure, and GCP with vertical-specific API frameworks for healthcare EHRs, real estate MLS systems, and finance platforms.

*Why did I build this?* Started with API keys, ended with self-evolving, self-healing Systems AGI... and I still have to retrieve my own API keys lol. At least the system is acquisition-ready, fully decoupled, heals its own memory leaks and stalled sockets natively, and scales horizontally.

Happy to answer any questions about the 13-layer self-healing mechanism, managing the websockets across a 12-server swarm without race conditions, or the multi-agent consensus algorithms. *Serious acquisition inquiries:* sophify-aaron@hotmail.com Tear it apart. What did I miss?

Comments URL: https://news.ycombinator.com/item?id=47230379

Points: 1

# Comments: 0

Article URL: https://bgthread.github.io/an-identity-crisis-in-engineering/

Comments URL: https://news.ycombinator.com/item?id=47230370

Points: 1

# Comments: 0

Article URL: https://mechanical-sympathy.blogspot.com/2013/02/cpu-cache-flushing-fallacy.html

Comments URL: https://news.ycombinator.com/item?id=47230363

Points: 1

# Comments: 0

Article URL: https://www.leadinginproduct.com/p/time-to-first-success-ttfs

Comments URL: https://news.ycombinator.com/item?id=47230361

Points: 1

# Comments: 0

I'm building a SaaS that accepts USDC payments. Simple requirement: customer sends stablecoins, I need to know when they arrived and the amount matches. Current options seem to be:

1. Payment processor (Coinbase Commerce, NOWPayments) — handles everything but takes 0.5-1% per transaction. On a $10k payment that's $50-100 just for confirmation.

2. DIY with Alchemy/Infura webhooks — poll for Transfer events, match amounts, track confirmations, handle reorgs, build retry logic. Took me ~2 weeks and it's still brittle.

Is there a middle ground? Something that just does the verification part — "tell me when X amount arrived at Y address" — without the custody/checkout/fee overhead? Curious how others are handling this, especially on L2s like Base where tx costs are negligible but verification infra is the same complexity as mainnet.

Comments URL: https://news.ycombinator.com/item?id=47230346

Points: 1

# Comments: 0

Article URL: https://jotter.jonathankingston.co.uk/blog/2026/03/03/stop-asking-junior-engineers-to-struggle-harder/

Comments URL: https://news.ycombinator.com/item?id=47230345

Points: 1

# Comments: 0

Article URL: https://github.com/OfficeDev/office-js/issues/6513

Comments URL: https://news.ycombinator.com/item?id=47230340

Points: 2

# Comments: 1

Article URL: https://techcrunch.com/2026/02/25/americans-now-listen-to-podcasts-more-often-than-talk-radio-study-shows/

Comments URL: https://news.ycombinator.com/item?id=47230335

Points: 1

# Comments: 0

Article URL: https://github.com/math-inc/Sphere-Packing-Lean

Comments URL: https://news.ycombinator.com/item?id=47230333

Points: 1

# Comments: 0

Article URL: https://twitter.com/mbrendan1/status/2028137693583221065

Comments URL: https://news.ycombinator.com/item?id=47230324

Points: 2

# Comments: 0

Article URL: https://arxiv.org/abs/2603.00278

Comments URL: https://news.ycombinator.com/item?id=47230315

Points: 1

# Comments: 0

Article URL: https://en.wikipedia.org/wiki/*Lisp

Comments URL: https://news.ycombinator.com/item?id=47230314

Points: 1

# Comments: 0