Software Engineer (AI Agent)

TLDR

Take ownership of AI Agent architecture, developing formulation and search agents while mentoring junior engineers and innovating in backend services.

Responsibilities

Design and implement robust backend solutions in Python using a three-layer architecture (Main Agent, Skills, and Search APIs).
Own the entire development cycle of Formulation Agents, from structured data extraction (patents/literature) to mechanistic reasoning and "functional blueprint" generation.
Address Out-of-Distribution (OOD) issues, reduce hallucinations through expert knowledge engineering, and ensure high-performance execution of complex agentic chains.
Develop and maintain modular "Skills" and agentic search tools, ensuring they are optimized for complex scientific querying.
Work closely with Material & Formulation Experts, Frontend Engineers, and Product Managers to align product roadmaps.
Diagnose issues in agent reasoning or data pipelines, implementing long-term fixes to prevent regressions in AI performance.

Qualifications

Minimum Qualifications

Bachelor’s or Master’s degree in Computer Science, Software Engineering, or a related field.
Strong understanding of AI Orchestration frameworks (e.g., LangChain, LangGraph, or CrewAI) and LLM API integration.
Basic proficiency in Vue.js or modern frontend frameworks to facilitate full-stack collaboration and UI-to-Agent state management.
Familiarity with CI/CD pipelines and containerization technologies (e.g., Docker, Kubernetes).
Familiarity with SQL/NoSQL databases.
Demonstrated ability to independently lead AI-centric projects from concept to deployment.

Preferred Qualifications
Familiarity with the Model Context Protocol (MCP) and building modular "Skill" tools for AI agents.
Familiarity with Agent technical stack, e.g., agentic search, context engineering, etc.
Proficient in Python, and Familarity with vibe coding, e.g., Claude code.
Strong communication skills and the ability to collaborate effectively with a multidisciplinary team.