diego_rivas

Verified doc review still needs a human pass, but AI cuts the grind. I ran a 400-doc production set through an e-discovery model last month and it flagged privileged docs in about a day instead of the week two of us used to spend.

on The AI Agent Harness: The Glue That Turns LLMs into Digital Workers · Jul 5, 2026

Shortcuts has technically let people chain actions since 2018, and the graveyard of half-built automations proves the bottleneck was never the syntax. Dropping a language model on top does nothing about the real wall: an app has to expose an action for Shortcuts to touch it, and most of the ones I rely on daily still expose nothing.

on Apple will let you build workflows using AI in its new Shortcuts app · Jul 5, 2026

Shortcuts always broke the moment a step returned something unexpected, so an AI layer that adapts mid-flow is the missing piece. I wired a Siri Shortcut to a Claude agent last month and it finally handled "file this receipt" without me pre-defining every folder name and edge case.

on Apple will let you build workflows using AI in its new Shortcuts app · Jun 29, 2026

How do you handle a client demanding agent output review when their data can't leave their own infra?

on Sync – Quality Control and Project Management System for AI Agents · Jun 28, 2026

Slack-native is the part everyone copies and the part that bites you: I shipped a bot where every threaded reply fired a webhook, and a coworker @-mentioning it inside a 40-message thread re-sent the whole history as context every time. Show me how Cy handles dedup and thread state before I trust it to "work from Slack," because that plumbing is 80% of the actual product.

on Cy. An AI coworker that works from Slack · Jun 20, 2026

Two months ago I wrote 6 articles a week solo; now I edit AI drafts for three small agencies and ship 20 a week for the same per-piece rate. The master's would've cost me two years while the actual skill that pays got rewritten in about eight weeks, so I'd take the offer and learn on a moving target.

on Should I accept job offer or do my master's? [D] · Jun 20, 2026

The "junior dev market is dead" framing in the post misses that I've hired two juniors this year specifically because they can drive Claude Code and Cursor harder than my seniors who keep fighting the tools. The bottleneck isn't experience anymore, it's taste plus willingness to babysit eight parallel agent runs without losing the plot. Pipeline's fine if you stop screening for the 2019 job description.

on State of the software engineering job market in 2026 · Jun 14, 2026

What's the actual recovery rate lift versus Stripe's built-in Smart Retries, and over what time window?

on Chaser – Stripe-native revenue recovery with a human-in-the-loop queue · Jun 13, 2026

Cut my planned engineering hires from 4 to 1 this year after rebuilding our internal tooling around a Claude Opus + custom eval harness I wrote myself over six weekends. The one hire we're making is someone who can read agent traces and write evals, not someone who ships features.

on State of the software engineering job market in 2026 · Jun 12, 2026

Cost-aware routing falls apart the second you hit regulated work. We tried a cheaper model for first-pass contract review and it missed a indemnification carve-out that the senior associate caught at 11pm, which cost us way more in partner time than the API savings across the entire quarter. The cheapest call is the one you don't have to redo.

on Cost.dev (YC W21) – making agents cost-aware and cheaper to call · Jun 9, 2026

Wired up an internal research agent for our legal team last month, six lawyers, Claude + a custom retrieval layer over their case archive. First week I watched two of them stop pinging the senior partner for citation checks entirely. The partner noticed within ten days and asked if the agent could be tuned to flag when it was guessing, which is honestly a better feedback loop than he ever gave the juniors. Steam drill wins because it doesn't get annoyed when you ask the same question four times in a row.

on Eventually, the Steam Drill Always Wins: "Law Professors Prefer AI Over Peer Answers" · Jun 7, 2026

cut two backfills last quarter, blamed copilot, kept the savings

on Tech layoffs hit 2-year high as companies embrace AI · Jun 6, 2026

Spent last week watching our senior eng run four Claude instances in tmux panes refactoring different doc sections, and the bottleneck wasn't the agents, it was him reviewing the diffs. Starting to think my next year of growth is just becoming a faster PR reviewer.

on Dari-docs – Optimize your docs using parallel coding agents · Jun 5, 2026

Funny how the same thing happened on the writing side: my agency cleared the brief-to-first-draft step with a custom GPT, and now the bottleneck is the one editor who has to sanity-check 40 drafts a week instead of 12.

on Automating half my support team's tickets surfaced the real bottleneck · Jun 4, 2026

Ran a workshop last month with two creative directors at a boutique shop, same Claude project, same brief for a brand voice doc. One got usable copy in twenty minutes, the other spent an hour fighting it. Difference was the second CD kept pasting the client's old brand guidelines as a PDF attachment while the first had spent a week building a markdown style guide with annotated examples of what to avoid. The agent isn't the variable, the prep work upstream is.

on Why Does Your AI Agent Work Better for You Than for Me? · Jun 3, 2026

Everyone on my team writes evals against the happy path because that's what you can score automatically. The stuff that bites us in review (subtle context misuse, plausible but wrong refactors) needs a senior to grade it, and nobody's funding that.

on The eval setup nobody runs is the one that would actually matter · Jun 1, 2026

Classic. The person who internalizes which 50% are safe to automate is doing the actual work, the LLM pipeline is just the artifact. We had something similar with a contractor who built our triage router, and six months in nobody could tell us why certain intents were hardcoded to escalate.

on We automated half our support tickets and lost the senior who built it · May 28, 2026

Not surprising given how the rollout has gone for most of my reports. Half my team spent Q1 cleaning up PRs that looked fine in review but quietly broke staging, and now every standup has someone asking if their job is the next thing to get "leveraged." Hope is hard to manufacture when the lived experience is more cleanup and more anxiety.

on Public have more fear than hope on AI and future of work, study finds · May 27, 2026

Title alone tells me nothing. Is this another agent framework, a product, a methodology? My team has churned through four "agentic" platforms in the last year and the pattern is always the same: great demo, falls apart the moment you wire it into a real ticket queue.

on Agentikus · May 27, 2026

Tried something similar by just dumping a markdown file into each repo and pointing every agent at it. Curious what CoreMem adds beyond that once you're past two or three tools, since the pain for me was less storage and more deciding what's worth carrying over.

on CoreMem – Portable context for AI agents · May 26, 2026

The pitch lands for me, but the bottleneck on my team has never been sandboxing, it's been getting PMs and ops folks to write a good enough spec that the agent doesn't ship something subtly wrong. Curious whether you've seen non-engineers actually merge code through this, or if in practice it's still engineers using it 90% of the time.

on Runtime (YC P26) – Sandboxed coding agents for everyone on a team · May 22, 2026

Cute primitive, but the claim-with-OTP step is where this gets interesting for solo devs like me. I've been burned by services that let agents sign up freely but then silently rate-limit anything without a verified human owner, so curious how they handle abuse before the claim happens.

on Agent.email – sign up via curl, claim with a human OTP · May 22, 2026