Gen AI Developer

• Strong proficiency in Python (async, typing, packaging, unit testing); solid grounding in data structures and algorithms
• Hands‑on with AWS—specifically Amazon Bedrock, SageMaker, S3, Lambda, Step Functions, CloudWatch, IAM; vector search - pgvector, OpenSearch vector
• Working knowledge of LangChain/LangGraph, RAG architectures, prompt design & evaluation; familiarity with guardrails (policy, PII/Secrets redaction), HITL patterns
• Experience with ETL/data processing (Pandas / Spark) and NoSQL/SQL stores (DynamoDB/Postgres)
• Proficiency with Git, CI/CD (GitHub Actions/Jenkins), Docker, and basic Kubernetes concepts
• Understanding of microservices and event‑driven integrations (queues, webhooks); observability (logging, metrics, tracing)
• Bonus: ReactJS for full‑stack prototypes; Copilot usage patterns for engineering productivity; document pipelines

Required qualifications to be successful in this role:

• What you must bring (Experience)

• 4–7 years of software engineering experience GenAi services focused, including 2+ years building or integrating GenAI/ML solutions
• Bachelor’s degree in Engineering/Computer Science (or equivalent)
• Practical understanding of IT + business concepts and how AI features translate to measurable outcomes
• Responsibilities – What will you do
• Design and develop AI‑powered backend systems and services that leverage Bedrock (model selection, orchestration), RAG pipelines (indexing, retrieval, evaluation), and S3 content sources
• Build secure APIs that integrate AI features into internal apps; partner with UI teams on full‑stack delivery
• Implement prompt engineering, evaluation harnesses, safety filters, and HITL review workflows; measure quality (precision, hallucination rate, latency, cost)
• Create ETL/ingestion jobs and embeddings pipelines; optimize chunking, metadata, and retrieval performance
• Productionize with CI/CD, IaC, logging/monitoring, and cost/latency optimization; contribute to runbooks and SLOs
• Collaborate with product owners, domain SMEs, data engineering, security, and compliance to align on value, controls, and go‑live readiness
• Champion reusability by packaging patterns (RAG, agents, evaluation) as internal accelerators and documenting best practices

• Nice‑to‑have / Good‑to‑have
• Experience with agentic workflows (task decomposition, tools, function‑calling) and evaluation frameworks
• Knowledge of SageMaker model endpoints, fine‑tuning/parameter‑efficient tuning, and feature stores
• Familiarity with pgvector/OpenSearch, prompt/test case libraries, cost observability, and Copilot productivity dashboards