AI Agent Engineering
• Build and own production-grade agent systems end to end, from solution design through deployment, steady-state maintenance, reliability improvements, and operational support.
• Implement stateful, durable agentic workflows with checkpoints, persistence, safe retries, escalation paths, and human-in-the-loop controls for high-impact actions.
• Design agent architectures that include planning, tool use, memory, orchestration, multi-agent collaboration, validation, and fallback mechanisms.
• Build secure tool integrations using MCP-based connectors, approved enterprise APIs, internal services, and tool registries with appropriate authentication, authorization, and auditability.
• Implement advanced retrieval and grounding patterns, including hybrid retrieval, vector plus structured retrieval, reranking, relevance tuning, context assembly, and answer validation.
• Treat evaluation as an engineering discipline by creating test datasets, regression gates, quality metrics, and release criteria for task success, groundedness, reasoning quality, and tool-call correctness.
• Instrument AgentOps and LLMOps practices by tracing full agent trajectories, including retrieval steps, model calls, tool calls, intermediate reasoning states, outputs, latency, cost, and failure patterns.
• Monitor production AI behavior and identify issues related to hallucination, tool misuse, state errors, loop behavior, poor retrieval quality, incorrect reasoning paths, and workflow failures.
• Apply AI safety, governance, privacy, risk, and auditability controls appropriate for enterprise-grade AI systems.
• Use modern AI engineering workflows and developer copilots responsibly to accelerate delivery while maintaining code quality, review discipline, and production standards.
• Stay current with fast-moving GenAI and Agentic AI tooling, frameworks, research patterns, and implementation practices, and translate relevant innovations into pragmatic engineering solutions.
Backend Engineering
• Build robust backend services and production APIs that power AI and agentic applications, primarily using Python, FastAPI, Pydantic, and modern service design practices.
• Design and implement clean, modular, testable, and maintainable backend components using strong software engineering fundamentals, type hints, async patterns, and SOLID principles.
• Develop microservices, event-driven workflows, job execution patterns, and integration layers using technologies such as Kafka, workflow engines, and cloud-native services where appropriate.
• Integrate with SQL, NoSQL, and vector databases such as PostgreSQL, MongoDB, Chroma, Pinecone, Qdrant, Weaviate, FAISS, or similar platforms.
• Build reliable data and retrieval pipelines that support embeddings, chunking, indexing, metadata filtering, structured retrieval, and downstream AI workflows.
• Deploy and operate services in real environments using Docker, Kubernetes, CI/CD pipelines, cloud platforms such as AWS, Azure, or GCP, and infrastructure best practices.
• Optimize systems for cost, latency, throughput, reliability, scalability, and quality using practical routing, caching, batching, model/provider selection, and performance profiling.
• Implement production observability for APIs, agents, jobs, workflows, and integrations, including logs, traces, metrics, alerts, dashboards, and runbooks.
• Own production troubleshooting for assigned components, perform root-cause analysis, and implement long-term corrective actions to improve reliability.
• Production deployment and support experience is mandatory. Candidates should have experience building, deploying, monitoring, troubleshooting, and maintaining applications actively used by business users in production environments.
Delivery and Collaboration
• Partner with business stakeholders, product owners, delivery teams, and technical teams to translate problem statements into practical, scalable, and production-ready AI solutions.
• Collaborate with architects, senior engineers, and technical leaders on solution design, architecture decisions, implementation approach, and delivery planning.
• Contribute implementation-focused recommendations during technical design discussions, while aligning with broader platform direction and architectural standards.
• Own assigned workstreams independently and deliver high-quality components that integrate cleanly into broader enterprise solutions.
• Make sound technical decisions by evaluating business impact, operational risk, quality, cost, latency, security, maintainability, and delivery timelines.
• Work closely with data science, backend, platform, QA, DevOps, and client delivery teams to ensure solutions are engineered, tested, deployed, and supported effectively.
• Contribute to engineering best practices, documentation, runbooks, reusable components, and platform accelerators that improve delivery consistency.
• Share knowledge, participate in code reviews, and support junior engineers when needed, without requiring people-management or roadmap ownership responsibilities.
• Participate in a culture of applied innovation where strong ideas are captured as patents, whitepapers, technical write-ups, demos, and reusable accelerators.
Qualifications and experience we consider to be essential for the role:
• Typically 5 to 8 years of professional experience in software engineering, applied AI, machine learning engineering, data engineering, or related disciplines.
• Demonstrated hands-on experience building, deploying, and supporting production-grade AI, GenAI, or software solutions that support critical business operations.
• Strong Python engineering experience with modern development practices, including testing, code reviews, debugging, version control, packaging, dependency management, and documentation.
• Experience building LLM-powered applications, including prompt design, prompt versioning, tool calling, RAG, evaluation, observability, and production readiness considerations.
• Practical experience with agentic frameworks or orchestration tools such as LangGraph, LangChain, AutoGen, CrewAI, LlamaIndex, or similar technologies.
• Hands-on experience building APIs, backend services, microservices, or integration services using frameworks such as FastAPI or comparable technologies.
• Working experience with SQL, NoSQL, and vector databases, along with data modeling, indexing, retrieval tuning, and performance considerations.
• Practical cloud deployment experience on at least one major cloud platform, with exposure to Docker, Kubernetes, CI/CD, monitoring, and operational readiness practices.
• Production deployment and support experience is mandatory, including monitoring, troubleshooting, incident analysis, and corrective action ownership.
• Strong logical reasoning, structured problem-solving, and ability to operate effectively when requirements are incomplete or ambiguous.
• Ability to evaluate business and technical trade-offs and make practical decisions that balance quality, speed, cost, scalability, risk, and user impact.
• Strong ownership mindset with experience driving assigned solutions from concept through implementation, deployment, production support, and continuous improvement.
Skills and Personal attributes we would like to have:
• Hands-on builder mindset with strong attention to code quality, system reliability, and maintainability.
• Strong collaboration skills and ability to work effectively with technical leaders, business stakeholders, and cross-functional delivery teams.
• Ability to communicate complex AI and engineering concepts clearly to technical and non-technical audiences.
• Strong debugging, root-cause analysis, and production support mindset.
• Comfort working in fast-paced environments where requirements may evolve and practical engineering judgment is required.
• Customer-focused and outcome-oriented approach, with the ability to connect technical decisions to measurable business value.
• Experience with knowledge graphs, structured retrieval, complex task decomposition, auto-labeling, synthetic data generation, or evaluation datasets is a plus.
• Exposure to developer copilots or rapid prototyping tools such as Cursor, Windsurf, GitHub Copilot, Claude Code, Codex, or similar tools is a plus.
• Prior experience supporting client-facing delivery teams or building reusable accelerators used across multiple engagements is a plus.