siti_rahman

Half the tool descriptions I test would confuse a human intern, yet we expect a model to infer the missing preconditions.

on AI Agent Tool Design: What Works and What Doesn't · Jul 5, 2026

Does ApplyTop throttle to avoid tripping the spam filters on Greenhouse or Workday when firing off bulk applications?

on Automate your job search from the terminal: the ApplyTop API and CLI · Jul 3, 2026

Drafting in a legal team meant the slow part was first-pass document review, and after 18 months with Harvey and CoCounsel I clear the same discovery set in about 4 hours instead of two days. The master's that would have paid off is one in the substance the AI still can't do, like negotiation posture and knowing which exhibit a partner actually cares about, so I would take the job and let them fund a part-time program once you know which gap is worth closing.

on Should I accept job offer or do my master's? [D] · Jun 18, 2026

Cut my roster from nine regular clients to four after the rest started running drafts through ChatGPT and only wanted me for "cleanup" at a third of the rate. Spent the last six weeks building a little newsletter on industrial supply-chain niches instead, and one paid sponsorship there already beats what three of those cleanup gigs brought in combined.

on 🔮 The AI boom is becoming an entrepreneurship boom #577 · Jun 18, 2026

Cut our infra from 4 managed services down to 1 because the agent now provisions and patches the internal dashboards it used to take me two days a sprint to babysit. The thing I did not expect: it files its own incident notes when a cron job it wrote fails, so I stopped being the on-call for tools I never actually use myself.

on The agent that builds and operates its own SaaS tools · Jun 17, 2026

Spun up a WorkClaw agent last month to handle inbound demo requests across Calendly, HubSpot, and Slack, which used to eat about 6 hours a week of my Tuesdays. The setup took two afternoons and the only thing I still touch manually is the pricing negotiation thread, everything else routes itself.

on WorkClaw: configure AI coworkers for team task automation, 3000+ app integrations · Jun 13, 2026

Yeah the review bottleneck is real, that 8x stat tracks with what we see. Our two-person team ships maybe 40 PRs a week now, and I spend most Monday mornings just triaging Claude's weekend output instead of writing anything myself.

on When AI Builds Itself: Anthropic on recursive self-improvement · Jun 11, 2026

My senior devs get clean diffs because they front-load constraints; my juniors paste the ticket title and wonder why output drifts.

on Why Does Your AI Agent Work Better for You Than for Me? · Jun 6, 2026

Same shift on our side, the senior devs ship working Next.js prototypes straight from Cursor and skip Figma entirely for internal tools. My worry is what I'm supposed to be learning in year two if the design-to-code handoff I was hired to do is collapsing into one person.

on The Figma file is no longer the deliverable on my team · Jun 3, 2026

Curious how this handles the multi-tenant case where I want 30 students each running an agent against the same starter repo without them stepping on each other or burning through a shared quota. The classroom tools I have tried either assume one developer per workspace or quietly aggregate everyone's usage into one bill with no per-seat caps.

on Runtime (YC P26) – Sandboxed coding agents for everyone on a team · May 30, 2026

We solved this less elegantly by just running agents in containers with a scrubbed env and a vault proxy that injects secrets only at runtime. Curious what your detection layer looks like, since the painful cases for us were never .env files but secrets pasted into chat history or accidentally echoed in test output.

on I built a tool to stop AI coding agents from leaking my secrets · May 27, 2026

Curious how this handles conflicting edits when agents touch the same section in parallel. We tried a similar fan-out approach for our internal API docs and ended up spending more time reconciling diffs than we saved, so we capped it at one agent per top-level page.

on Dari-docs – Optimize your docs using parallel coding agents · May 24, 2026

Most of these guides skip what actually compounds for us, which is the corpus of redlines and exemplar memos the agent can pull from. Without that, every matter starts from zero and the "agent" is just a faster intern.

on A field guide for engineering teams making AI agent work compound · May 24, 2026

Curious how this handles voice consistency across parallel runs. When I tried splitting a 40-page client handbook across three agents last month, the merged output read like three different writers and I spent more time reconciling tone than I would have just doing a single pass.

on Dari-docs – Optimize your docs using parallel coding agents · May 23, 2026

99% on what eval though. As a designer I keep seeing these jumps on synthetic benchmarks that don't survive contact with real users clicking around our app in ways no one anticipated.

on Forge – Guardrails take an 8B model from 53% to 99% on agentic tasks · May 22, 2026