A field methodology for hallucination, contradiction, and drift detection in memory-augmented agents, developed against a system in continuous production rather than a benchmark harness. github.com/rajaii-labs/agent-eval-methodology
Long-form technical work by Ramin Reza Rajaii on building and evaluating production LLM agents.
A field methodology for hallucination, contradiction, and drift detection in memory-augmented agents, developed against a system in continuous production rather than a benchmark harness. github.com/rajaii-labs/agent-eval-methodology
Granting autonomy to a production LLM agent, the specification it wrote for its own interior, and five silent defects found by following the money. github.com/rajaii-labs/agent-cost-forensics
Persistence, forgetting, self-models, and dreams — the buildable half of the question.
Peer-reviewed manuscripts, systematic reviews, and meta-analyses are listed on the Publications page. Regulatory writing includes eight R&D dossiers authored for biotech C-suite and investor audiences across six therapeutic areas, described on the Rajaii Labs page.