InsightRX | Engineering
About the Role
InsightRX is looking for a Full Stack Engineer to help build our Astra AI platform — a system that combines a modern web application with conversational AI agents to support pharmacokinetic/pharmacodynamic (PK/PD) analysis and clinical trial data workflows. You'll work across the client, server, and agent layers, understanding how they fit together end to end: how the frontend talks to the API, how the API queues and tracks work, and how LLM-powered agents process data and return results.
This is a strong fit for someone who enjoys working across the stack, is comfortable navigating both traditional web application architecture and newer LLM-agent design patterns, and wants to build tools used directly in pharmaceutical R&D decision-making.
What You'll Work On
- Build and maintain features across our React/TypeScript client and NestJS/TypeScript backend, including UI components, REST APIs, and the data layer connecting them.
- Design and extend the client-server architecture: state management, routing, authentication (including SSO/SAML), real-time updates via WebSockets, and background job processing via AWS SQS.
- Collaborate on the integration layer between the web app and our LLM-based agents — message schemas, queue-based communication, and file/data handoff.
- Work with our conversational AI agent system, which orchestrates multiple specialized agents for statistical modeling, data quality checks, and PK/PD analysis workflows.
- Contribute to database schema design and migrations (PostgreSQL/Sequelize), API documentation (Swagger), and test coverage (Jest, pytest).
- Participate in code review, architecture discussions, and production support across the stack.
What We're Looking For
- Strong professional experience building production level full stack web applications, comfortable owning features from database to UI.
- Strong TypeScript skills across both frontend (React) and backend (Node.js/NestJS) contexts.
- Experience designing REST APIs, relational data models (PostGresSQL, MongoDB), and authentication/authorization flows.
- Familiarity with message-queue or event-driven architectures (SQS, similar).
- Comfort reading and reasoning about systems that integrate with LLMs — understanding agent/tool patterns, prompt-driven workflows, and the practical quirks of working with LLM output, even without deep AI/ML research background.
- Strong problem-solving, ownership, and ability to debug across web applications and nondeterministic AI behavior.
- Uses Claude Code, Codex, or similar AI coding tools effectively while understanding, validating, and taking responsibility for the resulting code.
- Good instincts for reliability and correctness in domains where data quality matters — you ask "what happens when this fails or the data is malformed."
- Clear communicator, comfortable working across an evolving codebase with detailed technical documentation.
- Operate with a high level of independence and cope with ambiguity
Nice to Have
- Experience with LLM agent architecture a plus
- Experience with SAML SSO, multi-tenant systems, or healthcare/life-sciences software.
