Agentic AI Developer / AI/ML Engineer

Job Title: Agentic AI Developer / AI/ML Engineer

Location: Atlanta, GA

Duration: 12+ Months

Rate: DOE

Experience: 8–10+ years total (5+ years hands-on GenAI/Agentic AI)

Must-Have Skills

·         Python (expert level), backend services (Fast API/Django/Flask)

·         Agentic frameworks: Google ADK (Agent Development Kit), LangGraph, Lang Chain, MCP (Model Context Protocol), multi-agent orchestration

·         RAG architecture: embeddings, chunking, semantic/hybrid retrieval, reranking

·         Vector databases: Pinecone, FAISS, Weaviate

·         Agent evaluation: task success, tool-call accuracy, groundedness, latency/throughput regression testing

·         Prompt engineering & prompt versioning (templates, evaluation gates, rollback)

·         Cloud: Google Cloud Platform Vertex AI, BigQuery, GKE (AWS Bedrock/SageMaker or Azure AI a plus)

·         Observability: Lang Smith, MLflow, Prometheus, Grafana, ELK

Nice-to-Have

·         Agent/Tool Registry design (governance, versioning, access control)

·         Agent FinOps (token/cost attribution, usage analytics)

·         AutoGen, CrewAI

·         OAuth2/OIDC, RBAC/ABAC for agent-tool auth

·         ML frameworks: Scikit-learn, TensorFlow, PyTorch, MLOps (Docker, Kubernetes, Terraform, CI/CD)

·         Kafka/event-driven pipelines

·         Domain exposure: financial services, healthcare, or other regulated enterprise environments

Responsibilities

·         Architect agent workflows: planning, reasoning, tool calling, memory/context, guardrails, human-in-the-loop

·         Build reusable Python/TypeScript SDKs, REST/GraphQL endpoints for agent tools

·         Own RAG pipeline design and tuning for relevance, grounding, latency

·         Build agent evaluation harnesses and observability dashboards

·         Implement secure, auditable agent-to-system integrations (MCP, JSON-RPC, OAuth/RBAC)

·         Partner with architects, security, DevOps, and business stakeholders on production readiness