2,501 results for process · 2.205s

News for “process”
20 results • 2200 ms server time
Moozonian News
news.ycombinator.com• Apr 19, 2026• 1 min read
What early exponential growth can't tell us about runaway processesThis is an essay about a simple but (I think) underappreciated point:If a process follows a logistic curve, then early behavior is almost indistinguishable from exponential growth, which means you generally can’t infer its eventual carrying capacity from early data.In other words, very different long-term outcomes can all be consistent with the same initial “runaway” phase, even with very good data.I use Project Hail Mary as a framing device (basically mathematical fanfiction), but the real issues are inference and model choice.Curious whether people think this is a fair critique, or whether I’m overstating how ill-conditioned such problems are in practice.https://mathenchant.wordpress.com/2026/04/18/sorry-mary
Advertisement
Moozonian News
news.ycombinator.com• Mar 25, 2026• 1 min read
Show HN: Herd – A Go sidecar to stop stateful processes Puppeteer/LLMs from OOMHey HN.I'm an engineering student at Waterloo building stateful AI agents, and I kept hitting the same wall: whenever my Python scripts crashed or dropped a connection, the underlying Puppeteer or Ollama processes would just sit there orphaned, eating RAM until the node OOM-killed itself. Standard load balancers break sticky sessions, and passive HTTP timeouts are too slow for cleanup.I couldn't find a good local process pool that actually cleaned up dead stateful sessions reliably, so I built Herd in Go.It uses a persistent stream (gRPC/Unix sockets) strictly as a dead-man's switch. If your client script dies, the stream breaks. Herd registers the EOF and instantly fires a SIGKILL to the worker process (relying on Pdeathsig on Linux). For the actual heavy data, you just blast HTTP traffic through Herd's internal proxy, which routes it directly to the active process port.My actual goal is to turn this into a multi-node distributed mesh with a Redis registry, where a client can drop off
Advertisement
Moozonian News
app.simplai.ai• Mar 18, 2026• 1 min read
Show HN: We built AI agents that reduce mortgage processing from 18 days to 3–5Most mortgage processing delays aren’t due to risk — they’re due to manual workflows.We’ve been working on SimplAI, an AI-driven system designed for banking and financial services, starting with mortgage operations.The problem we kept seeing:15–22 day processing timelinesHeavy manual document handling (500+ pages per loan)Repetitive data entry + verification loopsUnderwriters spending hours on non-decision workSo we built a set of AI agents that handle the operational layer:Document AI (IDP) → classifies + extracts data from loan docs in minutesIncome analysis models → parse tax returns, payslips, and variable incomeVerification integrations → real-time employment + financial checksAI-assisted underwriting → pre-validates files and generates conditionsCompliance engine → continuously checks against regulatory rulesWhat we’re seeing in production:End-to-end processing: ~18 days → 3–5 daysData extraction accuracy: 97%+Underwriting review time: 3–4 hrs →
Advertisement