Antonio Monteiro
@_anmonteiro
Eng @recallai, building the universal API for realtime meeting data. Making @melange_re
I am now on Github Sponsors. If you appreciate my work and want to support the projects that I work on publicly, you can now show your generosity by supporting me directly! github.com/sponsors/anmon…
We hit the linux file descriptor limit in AWS RDS Postgres, turns out there’s a hard cap of 6815744. We had to get an AWS SDE to SSH into the actual VM running the RDS instance and execute some bash commands to raise the kernel-level FD limit. This is a totally undocumented…
New reason-react version released today (0.16.0) Support for @reactjs 19.1 (use, useActionState, useOptimistic), stable useTransition, integration with RSC, documentation updates, Melange 5.x, and more! Work by me (@davesnx) @_tatchi @_anmonteiro and @__r17x
now at the top of the orange site 🫠
you really only learn about stuff like this when you run it at scale
you really only learn about stuff like this when you run it at scale
At scale, Postgres' `LISTEN/NOTIFY` adds a global mutex on `COMMIT`. We hit this at @recallai while handling tens of thousands of concurrent writers. It effectively serialized every `COMMIT`, brought the DB to a halt, and is barely documented. We only found it by digging…
At scale, Postgres' `LISTEN/NOTIFY` adds a global mutex on `COMMIT`. We hit this at @recallai while handling tens of thousands of concurrent writers. It effectively serialized every `COMMIT`, brought the DB to a halt, and is barely documented. We only found it by digging…
Great news everyone! FUN OCaml 2025 is happening in Warsaw, September 15 + 16, 2025! The CFP and registration are open 🧡🐫👇
we're hiring strong engineers in SF to work on really impactful infra / product. we're building the infrastructure to access conversation data in real time, operating at a huge scale with a tiny engineering team. does this sound interesting? DMs open
We ran 7 million EC2 instances last month. If this level of scale sounds interesting, please reach out 🙏
I havent snooped around but this sounds very exciting. I wonder what theyre using to orchestrate this, what kind of auto scaling theyre doing, the observability on this scale must be insane. I bet the healthchecks are super tight, and the persistence layer at this scale. sheesh!!
We ran 7 million EC2 instances last month. If this level of scale sounds interesting, please reach out 🙏
We ran 7 million EC2 instances last month. If this level of scale sounds interesting, please reach out 🙏
.@recallai has been growing really fast and we’re looking for engineers to help build the cloud for conversation data. We run something like ~40k CPUs at peak load and record millions of meetings every month. Check link below 👇
.@recallai has been growing really fast and we’re looking for engineers to help build the cloud for conversation data. We run something like ~40k CPUs at peak load and record millions of meetings every month. Check link below 👇
ahrefs frontend stack: - Melange - Reason - reason-react - react.js (react dates, recharts, etc...) - styled-ppx - server-reason-react - atd (migrating to melange-json) part-of-frontend-that's-actually-backend: - OCaml - Lwt - Routes - sqlgg - devkit (stdlib and webserver)