Senior AI Agent Engineer

Company Overview

Not specified.

Job Summary

We are seeking an experienced engineer to design and build secure, production-grade AI agents that seamlessly integrate with business systems. The role involves developing agents capable of retrieving trusted information, utilizing tools, and supporting human-approved workflows. The successful candidate will contribute to advancing AI capabilities within the organization, ensuring robust, safe, and efficient AI operations in production environments.

Responsibilities

• Build and operate the AI-agent runtime, ensuring reliable and scalable performance.
• Design coordinator and specialist-agent workflows to optimize task execution and collaboration.
• Create typed and versioned tool registries for consistent and maintainable integrations.
• Integrate agents with backend services and data sources to enable comprehensive data access.
• Implement conversation state management and purpose-bound memory to support contextual interactions.
• Develop Retrieval-Augmented Generation (RAG) ingestion, retrieval, provenance, and citation systems.
• Implement deterministic authorization and policy controls outside the AI models to enforce security and compliance.
• Incorporate human confirmation steps for actions with significant impact.
• Build evaluation suites and automated release gates to ensure quality and safety.
• Establish observability, latency, and cost controls for operational efficiency.
• Implement feature flags, kill switches, and graceful degradation mechanisms.
• Support security reviews, production deployment, and incident response procedures.
• Own canary releases, rollback procedures, and maintain operational runbooks for smooth deployment and maintenance.

Qualifications

• Strong Python engineering experience with expertise in async Python, FastAPI, Pydantic, SQLAlchemy, and Alembic.
• Proven production experience with LangGraph, OpenAI Agents SDK, or comparable agent frameworks.
• Deep understanding of Retrieval-Augmented Generation (RAG), embeddings, vector databases (e.g., pgvector), and source-grounded generation.
• Hands-on experience with PostgreSQL, Redis, and queue-based asynchronous background processing.
• Knowledge of REST APIs, Server-Sent Events, and distributed system fundamentals such as idempotency, retries, and timeouts.
• Experience with model-provider integration platforms like OpenAI, AWS Bedrock.
• Skills in model routing, fallbacks, context management, and prompt versioning.
• Ability to manage token, latency, and cost controls effectively.
• Proficiency in OAuth2, JWT, JWKS, and service-to-service authentication.
• Expertise in PII minimization, encryption, secrets management, and least-privilege access.
• Experience with Docker and cloud deployments, preferably on AWS.
• Familiarity with OpenTelemetry, structured logging, metrics, tracing, and alerting.
• Strong background in automated testing using pytest and managing integration-test environments.

Preferred Skills

• Experience in AI safety and evaluation, including regression testing, deterministic evaluation, and prompt-injection testing.
• Familiarity with tools such as DeepEval, RAGAS, Promptfoo or equivalent.
• Knowledge of adversarial and multilingual testing.
• Experience in regulated or high-trust domains is advantageous.
• Ability to design evaluation datasets and deployment gates.
• Understanding that model tool selection is distinct from authorization.

Experience

• 7+ years of software engineering experience.
• 2+ years building generative AI or large language model (LLM) systems.
• Proven track record of deploying tool-using or RAG-based AI applications to production.
• Strong background in backend and distributed systems fundamentals.
• Experience in moving AI products from prototype to monitored production environments.

Environment

• The role involves working in a collaborative, fast-paced environment.
• Likely involves remote, in-office, or hybrid work settings.
• Requires adherence to security protocols and incident response procedures.
• Focus on high-trust, regulated domains may involve additional compliance and security considerations.

Salary

Not specified.

GrowthOpportunities

Not specified.

Benefits

Not specified.