hey PH 👋 I'm Razvan. every team I've worked on has one unofficial job nobody's hired into: the translator. support pings, "what happened to order #4421?" and someone opens five tabs. the app logs, the background jobs, the database, a few greps. 20 minutes later they paste back one sentence: "address validation failed, the retry ran 3 times, then it rolled back." the customer gets an answer. the engineer lost their afternoon. that person is on every team, and it's usually your best engineer. i got tired of being mine, so i built the thing that does the translating. it's live today. 🧩 what it does drop one agent into your app and it captures everything that happens (every request, every background job, every database change) and folds it into one plain-english card anyone can read: "Maria placed an order for 3 books totaling €47.20. Payment cleared on the first attempt, inventory was decremented, the confirmation email was queued, and the warehouse was notified." same events your server already logged, just told the way a human needs them. the hard part isn't the wording, it's the correlation: stitching the job that ran 4 seconds after the request back to that request, resolving an internal id back to "User 42 (Razvan)". that's the real work, and it has to be right. 💬 ask your product anything, then let it watch pin a question the way you'd say it out loud. "tell me when a trial account goes quiet for 3 days." "is checkout failing for anyone?" it watches the answer for you and pings you the moment it changes. no metric to name, no query to write, no dashboard to build first. the founder, support, and sales each pin their own question and finally stop interrupting the one engineer who knows where to look. 🧾 every number shows its receipts (and there's no AI guessing) the translation is deterministic. no model in it, just correlation plus templates, same events in, same sentence out, every time. so every number clicks straight back to the real events behind it. "3 checkouts failed today" opens those exact 3 checkouts. nothing to invent, so nothing invented. (and yes, your AI can read the same log over MCP, with citations, never making rows up.) ⚙️ one line to install, more stacks coming bundle add ez_logs_agent for rails, npm i ezlogs-nextjs for next.js. restart, done, no per-route code. rails and next.js today, python/django and node/express next. and it's read-only by design: it never writes back into your app, so the blast radius is tiny. 🚫 what it's not: not datadog, not an APM, not a metrics tool. it doesn't watch CPU or page you at 3am. it explains the work your app did, in language a support person, a PM, or a founder can actually read. free tier, no card: https://ezlogs.io if you've ever been the translator on your team, i'd love your honest take. what's the one question you keep getting asked about your own product? Razvan

OtariControl Your LLM Infrastructure

Promoted

Maker

📌

6d ago

One thing I'd love to see is Slack or Teams integration so the team can get pinged in the channels they already live in rather than checking another dashboard. Maybe even let non-engineers subscribe to specific activity types so product or support folks only hear what matters to them.

23h ago

@brahimo9zc thanks for the comment. yes, that's the next feature I plan to add.

22h ago

finally something that explains our background job failures without me digging through 4 dashboards. the click-back to raw events is honestly what sold me, no more guessing why a number shifted

1d ago

@dne3v07 thanks for checking it. I'm glad you like it.

EZLogs

Ask your product what happened, in plain english

Ask your product what happened, in plain english