Building production agent infrastructure & constrained-decoding systems.
I measure what LLM systems actually need — context, model capacity, harness — at constant quality.
Now shipping idfagent — multi-harness, sandboxed agent runtime in production.
I build production agent infrastructure and constrained-decoding systems for LLMs. 15+ years shipping production AI — agent runtimes, grammar-constrained decoding, neuro-symbolic agents, distributed systems. Trained through Iran's exceptional-talents system.
Currently architecting idfagent — a multi-harness, multi-provider, sandboxed agent runtime in production at evolutionID. The thesis: the session, not the harness, is the canonical compliance and evaluation unit.
Inventor of GRID (Grammar-Railed Decoding, EP 4687028) — patent deployed in production coding agents.
Co-founded and led two AI companies as CTO; now CTO at evolutionID, building agent infrastructure hands-on. Three granted patents (EPO, USPTO) and two further applications.
Specializations: Agent Infrastructure • Constrained Decoding • Test-Time Learning.
Munich-based; open to building in San Francisco.
Directed a research roadmap yielding 6 peer-reviewed publications (LREC, SPIE) that directly translated into enterprise product capabilities. Spearheaded the creation of MirasText, the largest automatically generated Persian corpus, to unlock high-accuracy, at-scale Sentiment Analysis, NER, and Topic Modeling for unstructured social streams. Continuously drove platform innovation by deploying Active Learning to slash annotation costs, engineering graph-based algorithms for trend extraction in noisy data, and developing multimodal audio systems (MirasVoice, Acoustic Fingerprinting).
Service: Invited reviewer, SURGeLLM workshop @ ACL 2026.