Skip to content
View rogermsc's full-sized avatar

Highlights

  • Pro

Block or report rogermsc

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
rogermsc/README.md

Roger Simões

Applied AI engineer. I build AI that works in the real world.

Computer Scientist and Mechanical Engineer with 20+ years shipping technology across three continents. I build LLM and agent systems, computer vision, and applied AI that runs in production: on the factory floor, the warehouse, the edge device, not just in a chat window. I founded and exited a computer-vision company, and I still stay close to the code.

make AI useful, not cute

Python PyTorch OpenCV Docker PostgreSQL Linux Git


What I build

  • LLMs & agents · agentic workflows, tool use, memory, evaluation, RAG, and prompt/context engineering
  • Computer vision · real-time detection and analytics on live camera streams
  • Applied & industrial AI · predictive maintenance, automation, and decision systems for heavy industry
  • Edge AI · self-improving agents designed to run on-device

Featured project

  • darwin-memo · self-curating memory for LLM agents, kept honest by survival-based selection rather than reward models or judges. Python, published on PyPI, with an MCP server, a benchmark suite against baselines, and integrations for the OpenAI Agents SDK, Anthropic, and local Ollama.

Currently

Leading applied-AI delivery for industrial clients (Timberstone) and product for an LLM platform at the Straits Institute for Applied AI. Author of two books on AI and the future of work.

Find me

Pinned Loading

  1. darwin-memo darwin-memo Public

    Self-curating memory for LLM agents: MeMo-style external memory kept honest by survival-based selection. No reward models, no judges.

    Python

  2. openclaw-memory-darwin openclaw-memory-darwin Public

    OpenClaw memory-slot plugin backed by darwin-memo: lessons earn or lose energy from measured run outcomes, no curation

    TypeScript