Engineer, Principal

Overview

Lead the design and delivery of production-grade agentic AI systems for enterprise workflows. Own agent architecture, knowledge systems, evaluation frameworks, and technical standards while partnering with domain experts and mentoring engineers.

What you'll do

• Design LangGraph-based agent graphs for multi-step workflows.
• Build MCP tool wrappers for enterprise data and domain tooling.
• Design ontology schemas and maintain RAG pipelines and vector stores.
• Build evaluation frameworks, golden datasets, and regression gates for agent outputs.
• Partner with domain experts to map workflows and define agent archetypes.
• Define agent scaffolds, ontology templates, and co-design standards.
• Mentor engineers and drive agentic-layer architecture decisions.

What you'll need

• 10+ years of software engineering experience.
• At least 3 years focused on AI/ML systems in production.
• Experience building production agentic or multi-step LLM systems.
• Strong Python skills, including async programming and type annotations.
• Experience with LangGraph or an equivalent graph-based agent framework.
• Experience with vector databases and embedding pipelines.
• Bachelor's degree in Computer Science, Engineering, Information Systems, or a related field.

Nice to have

• Experience with Temporal, Airflow, or equivalent workflow orchestration engines.
• Familiarity with OAuth, rate limiting, retries, and token management.
• Experience evaluating LLM outputs with domain-expert curation.
• Fine-tuning experience on domain-specific data.
• Master's degree with an ML or Systems specialization.

Details

• Location: Bangalore, India.