The methodology behind Observed Markets works for any business
Observed Markets is my public lab: a daily, fully transparent experiment in how far an agentic AI pipeline can go when exposed to noisy, high-variance data. If you want something similar running inside your company, I can help build it.
What I build
Process automation
Manual workflows turned into reliable, observable pipelines. Clear guardrails, audit trails, human-in-the-loop where it matters.
Agentic AI systems
Autonomous agents that plan, call tools, and make decisions. Not demos: systems that run daily and you can audit every step.
Data-heavy analysis
Pipelines that ingest messy, high-variance data and produce something you can actually act on. Same playbook as this site.
Rapid prototyping
From idea to working prototype in days, not quarters. Small, instrumented, easy to kill if it does not pan out.
How I work
- Small and senior. You get a single point of contact who writes the code, not a delivery pyramid. Fewer meetings, more working software.
- Outcomes over theatre. Short engagements with a measurable deliverable. No multi-year transformation programmes.
- Transparent by default. You see the code, the prompts, the evals, and every choice I made along the way. This site is proof of how I operate.
- Honest about limits. If an LLM is the wrong tool for your problem, I will tell you. The research on this site exists partly to map where LLMs fail.
Who I am
I am Ricardo Almeida Silva, based in Portugal. I build agentic AI systems for a living and run Observed Markets as a personal research project on the side. You can see how I think, how I write code, and how I handle uncertainty by browsing the research history on this site. If that resonates, get in touch.
Ready to talk?
Send a short email describing the problem you want to solve and the constraints you are working inside. I usually reply within a few days.
Want to see the work first? Browse the research history or read the methodology.