Overview
Rogerio Chaves is a co-founder of Langwatch, a company he started in October 2023. He is based in Amsterdam and states expertise in large language models and machine learning.
He joined Booking.com in July 2018 as a fullstack developer in FinTech, became a senior fullstack developer in the same area in June 2021, and from July 2022 to January 2024 was an engineering manager for Attractions. Before that he was at ThoughtWorks from January 2015 to May 2018, first as a consultant developer and then as a senior consultant developer. Between January 2011 and January 2015 he was the founder of React, and in early 2010 he held a web developer internship at AM4 - A Internet de Resultados.
Chaves studied at ETPC - Escola Técnica Pandiá Calógeras, qualifying as a computer technician between 2008 and 2010, and took a bachelor's degree in information systems at Centro Universitário de Volta Redonda - UniFOA from 2011 to 2014. He also studied at Sebrae in 2013.
Career history
In the news
- Introducing Instant Evals on LangWatch Run any eval you can think of, on all your production trace history. Cheap, fast, and at scale. Powered by TypeSafe AI's new Jev model. Your production data is gold, and now you can mine all of it as fast as your ideas. You can use it to find all sessions where users were frustrated, or where an extra tool call would have saved tokens, or to find what top questions your users ask You can also use it to improve your own coding harness, for example processing all your past Claude Code
- This is a great post, Prosus has always been at the forefront of enterprise AI adoption, exciting to see their journey to optimizing AI usage guided via evals! At LangWatch we have been helping customers with the same movement, if you are looking to optimize your AI costs, with routing guided by evals, context size control and possibly even moving to open models, hit me up! https://lnkd.in/egYQ8-J8
- Last week I was in London talking to dozens of founders and VCs, taking the tube around the city and rushing from a place to another, but totally worth it, incredibly exciting ecosystem, here is my overall summary: Last year Vertical AI SaaS was the hottest game in town, however this time across the board the excitement is going much deeper: model builders, AI infrastructure, horizontal platforms, hardware, robotics, even space and some genuinely difficult deep tech. Basically, much less fear of attacking really hard technical
- There are four kinds of AI agents running in a company: 1. The Custom Ones: built for a product feature or an internal process. Your flagship agent or an internal RAG, using LangGraph, Vercel AI, custom code on Azure OpenAI or Bedrock. They get a project, an owner and a budget. 2. The Coding Ones: Claude Code, Codex, Cursor. Bought as seats, but they read code, open pull requests and run commands. 3. The Assistant Ones: Claude Cowork, ChatGPT Enterprise, Microsoft Copilot, rolled out to everyone. Approved once, then every team
- Can you list every AI agent running in your company right now? Not the approved ones. The running ones. The picture is the same almost everywhere: a few agents on Copilot Studio, Claude Cowork rolling out org-wide, a Databricks Genie here and there, all devs using Claude Code, Codex or Cursor, plus AI quietly switching itself on inside the SaaS you already pay for, on Jira, Linear, Notion... Most started as somebody's initiative, others simply became available without you noticing, and it's all now too much to control. Most
- Ask all your engineers to compact at a 450k context window, this is the single biggest thing you can do to save tokens right now tl;dr: we did an analysis with 2,451 agent sessions from our own team, and found that every 2x increase in context size let to 6x more token cost, even cached, with no evidence that earlier tokens were needed in later work. We could never justify going all the way to 1M context window For a couple months now we have been tracing our own Claude Code usage telemetry with LangWatch, so when in just a few
- We are live on Product Hunt! LangWatch is the most powerful way for tracking your Claude Code, Codex, Copilot and OpenCode usage, for engineering teams to improve their own flows and skills to optimize their time, quality, and tokens If LangWatch has ever helped your team, a comment on the thread today means a lot. Help us out on our Product Hunt Launch! 👇 https://lnkd.in/dzezbcFP
- Track also codex, copilot, opencode, all in the same place, optimize and compare across harnesses, just run: npx langwatch codex (or opencode, or gemini, or copilot) cc Yevhenii Budnyk
Related profiles
This page shows public professional information only, each fact cited. Is this you? send a correction, or ask for removal within 24 hours, no questions asked.