Apply appropriate guardrails, access controls, validation, and safety mechanisms to ensure agents operate within defined boundaries.
Work with SQL/NoSQL databases to manage structured and unstructured data pipelines feeding agent workflows.
Contribute to full-stack delivery, connecting LLM backends to modern frontends.
Operate within existing access-control, logging, and audit patterns for multi-tool agent systems; flag gaps to the architecture team rather than redesigning independently.
Required Skills (Must Have)
6+ years of professional software engineering experience, including at least 1–2 years hands-on building with LLMs or agentic frameworks.
Strong proficiency in Python and/or Node.js for AI and backend development.
Working knowledge of major LLM providers (e.g., OpenAI, Anthropic) and prompt orchestration frameworks (LangChain, LangGraph).
Hands-on expertise in building LLM applications, including
Prompt engineering and enforcing structured outputs
Tool integration and function calling
Retrieval-Augmented Generation (RAG) using vector databases and embedding-based search
Designing multi-agent workflows and orchestration frameworks
Managing context windows and conversational state
Experience integrating external tools and APIs via function calling or tool-registration mechanisms.
Experience integrating agents or tools with enterprise SaaS applications via REST APIs or connector frameworks.
Working knowledge of SQL and NoSQL databases and how they plug into AI pipelines.
Experience writing test cases or evaluation harnesses for agent/LLM behavior (accuracy, hallucination, tool-call correctness).
Familiarity with agent evaluation techniques, including LLM-as-a-judge, task-based evaluations, benchmark design, and automated regression testing.
Understanding of access-control and audit-logging patterns in multi-tool, multi-agent systems.
Understanding of AI security, prompt injection, data leakage, permission boundaries, and agentic-system guardrails.
Strong debugging skills and the ability to implement complex multi-step systems within an established architecture.
Nice to Have Skills
Understanding of cloud deployment and infrastructure (AWS, GCP).
Knowledge of containerization and CI/CD pipelines (Docker, GitHub Actions).
Experience with AI observability, guardrails, and safety mechanisms for LLMs.
Familiarity with data warehouses or analytical query engines for large-scale structured retrieval.
