Roei Shemesh

Academic background

B.Sc. Electrical Engineering, Technion — Israel Institute of Technology (2013–2017), graduated with honors (GPA 90)

• Technion EE — systems, signals, and rigorous problem-solving; graduated with honors
• Machine Learning Developer at CEVA (2018–2022) — algorithms in a DSP / signal-processing IP environment
• Financial algorithms & kernels — market data → features, signals, and executable decisions under latency, noise, and risk
• LLM / agent systems — Baby Agent (local fine-tunes), KYA (masks), ReAct eval loops, Nexus (coordinated crew over shared memory)

Where I am now

I am actively looking for roles. For years I worked alone — shipping hard, sharing little. This site is the first deliberate step back into the open: still here, ready to collaborate again.

How I work

• Behavior as design: mature agents run on defined qualities, absolute level-0 rules, and relative rules for cooperating on a shared project.
• Governance: agent learning runs through a traceable pipeline with redaction and human approval — never raw logs.
• Making the abstract tangible: concepts modeled as physical systems; agents you can watch, inspect, and steer.
• AI as leverage in production — not a buzzword. Python since 2018.

See the coverage board for how LinkedIn agent lanes map to stories — and contact when you want to talk.