Note: The job is a remote job and is open to candidates in USA. InsightRX is looking for a Full Stack Engineer to help build their Astra AI platform, which supports pharmacokinetic/pharmacodynamic analysis and clinical trial data workflows. The role involves working across the client, server, and agent layers to develop features and maintain the system architecture.
Responsibilities
- 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
Skills
- 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
- Experience with LLM agent architecture a plus
- Experience with SAML SSO, multi-tenant systems, or healthcare/life-sciences software
Company Overview
Company H1B Sponsorship