GenAI Architect (Hands-On) — AI-Accelerated SDLC (Claude Code)
Roles and Responsibilities:
- Translate the customer's AI-driven SDLC roadmap into a clear target architecture, detailed design, and phased implementation plan.
- Design and implement the agentic framework and agent orchestration layer for SDLC use-cases using LangGraph (preferred) and LangChain where appropriate.
- Define the right approach per use case: tool calls vs. MCP servers vs. agents, including routing/orchestration patterns and multi-agent collaboration.
- Work with a small team (2 members) to build, test, deploy, and operate a portfolio of MCP servers, tools, and agents.
- Implement robust LLM integration patterns: prompt/agent design, context & memory management, guardrails, fallback strategies, and cost/latency optimization.
- Establish evaluation and quality gates for AI in the SDLC: automated tests, regression evals, golden datasets, and CI integration.
- Set up observability & operations: logging, tracing, prompt/version tracking, usage analytics, and runbooks.
- Produce enablement assets for Claude Code adoption:
- claude.md, memory.md, agent/tool conventions, repo templates
- How-to guides for agents, hooks, workflows, and team best practices
- Partner with engineering/security on data privacy, secrets management, and policy compliance.
Required experience
- 10+ years in software engineering / IT with strong architecture + hands-on delivery.
- 2+ years building with GenAI/LLMs in production (agents, tool use, RAG, orchestration).
- Strong LangGraph experience (state machines, multi-agent flows, human-in-the-loop), plus LangChain for tooling/integrations.
- Demonstrated experience adopting Claude Code in advanced usage (agents, hooks, workflows, repo conventions) and coaching teams.
- Strong engineering fundamentals: APIs, CI/CD, testing strategy, SDLC tooling, cloud basics.
- Working knowledge of classical ML concepts (certification a plus, not required).
Nice to have
- Experience building/operating MCP servers and tool ecosystems.
- Experience with enterprise security/compliance, data governance, and red teaming.
- Experience with developer productivity platforms and SDLC automation (GitHub/GitLab, Jira, code review automation).
Deliverables
- Target architecture + detailed design docs for AI-accelerated SDLC using Claude Code + LangGraph orchestration.
- Working MCP/tool/agent implementations with CI, evals, and monitoring.
- Playbooks/templates (claude.md, memory.md, orchestration patterns) and concise enablement documentation.
