Agentic AI Engineer (West Bengal)

Required Skills: Experience in backend/platform engineering with solid proficiency in Python/Java (or equivalent).
• Hands-on experience building workflow engines, orchestration layers, or distributed services with asynchronous execution patterns.
• Strong understanding of agentic AI: tool calling, planning loops, state management, memory strategies, and human-in-the-loop design.
• Experience integrating LLM workflows with enterprise systems (APIs, data sources, ticketing, workflow tools).
• Ability to engineer reliable systems: idempotent execution, retries, timeouts, circuit breakers, and structured logging/telemetry.
• Familiarity with secure-by-design practices: authN/authZ, secrets management, data boundary controls, and auditability. Preferred Skills: Design and implement agent workflows for manufacturing processes (e.g., incident triage, quality deviations, supplier risk signals, engineering change support).
• Build a robust orchestration layer that combines LLM reasoning + deterministic tools (APIs, rules engines, data retrieval, workflow actions).
• Implement guardrails: human approvals, policy checks, validation steps, and traceable decision logs for enterprise audit readiness.
• Create reusable patterns: agent templates, tool libraries, prompt/task decomposition approaches, and standard evaluation hooks.
• Ensure production readiness: monitoring, failure handling, incident playbooks, performance tuning, and cost controls.
• Collaborate with product owners, domain SMEs, and platform teams to industrialize pilots into repeatable AI services. Apply on Kit Job: kitjob.in/job/4m0yq1