Synthetic Brain System Engineer

Pure neural models are prone to hallucination; pure rule systems lack generalization. As a Synthetic Brain System Engineer, you will unite these paradigms into a robust dual-layer cognitive architecture where deterministic logic scaffolds guide neural generation and verify every computational state.

• Construct hybrid neurosymbolic processing pipelines that map natural language into deterministic state-machine transitions.
• Build structured grammar-constrained decoding engines that guarantee 100% schema-compliant JSON/SQL execution.
• Integrate probabilistic graphical models with neural embeddings for Bayesian risk and anomaly scoring.
• Develop automated rule-compilers that transform enterprise policy documents into executable symbolic verification guards.
• Continuously evaluate and benchmark neurosymbolic accuracy against standalone frontier LLM baselines.

• Formal language theory, context-free grammars, Pushdown Automata, and regular expressions.
• Neurosymbolic computation, knowledge representation, and ontology engineering.
• Probabilistic reasoning, Bayesian networks, and causal inference.

• Strong programming capability in Python and/or Rust.
• Experience building custom grammars using ANTLR, Lark, or regex state automata.
• Proficiency in LLM grammar-guided sampling libraries (Outlines, Guidance, LMQL).

• Passionate interest in combining symbolic logic with modern deep learning representations.
• Commitment to deterministic correctness in safety-critical business processes.
• Curiosity for cognitive models of human language parsing and rule deduction.

• Annual package ₹12,00,000–₹20,00,000 with annual revision.
• Pioneering work in commercial neurosymbolic architecture with patent potential.
• Work from anywhere in India with complete schedule flexibility.
• Full hardware and cloud computing allowances.