Ali Sohrabkhani

Lead Software Engineer and Global Banking and AI Champion with 20+ years of progressive experience designing solutions for satellite-data pipelines, multi-trillion-dollar bond and loan rails, and enterprise AI systems. Specialized in architecting state-machine agents that blend vector search with knowledge graphs, evaluating AI coding tools, and leading cross-functional teams building AI powered financial data ingestion and delivery pipelines. Proven track record of integrating LLMs into cloud-native applications, creating AI agent-driven workflows, and implementing RAG systems from structured and unstructured data. Committed to elevating developer productivity through intelligent automation so teams can focus on high-impact challenges while delivering faster, more accurate outcomes.

Current Focus Areas:
• LangGraph & Graph‑RAG Workflows: Architecting state‑machine agents that blend vector search with knowledge graphs to deliver grounded, traceable answers for both internal tooling and customer-facing applications.
• Developer Productivity: Building custom MCP servers that allow engineers to chat directly with microservices, internal APIs, and private knowledge bases from within their IDE, eliminating context-switching and boosting flow.
• LLM Benchmark Lab: Operating a continuous evaluation platform that compares Copilot, Codeium, DevGPT, Cline, Cursor, and Windsurf on latency, context handling, and code quality metrics to guide tool selection, licensing, and performance SLAs.
• Prompt Engineering and Training: Facilitating live workshops on advanced prompting techniques, schema-driven patterns, and artifact-based retrieval augmentation to embed best practices into daily developer workflows.

Highlights:
• Co-led development of an AI-driven code-refactoring engine that autogenerated microservice skeletons, cutting build and test cycles by 50%.
• Architected a 10 PB/day satellite telemetry ingestion pipeline with sub‑second query performance for mission-critical analytics.
• Served as AI Champion in the AI4Tech group, making strategic decisions on AI tooling, productivity assistance, code generation, and test automation.
• Founded and scaled an internal Gen‑AI guild, mentoring over 150 engineers across APAC and North America in best practices for LLM deployments.
• Orchestrated migration from a monolithic codebase to a poly-repo architecture, introducing code-generation templates that reduced new-service onboarding from one week to a few hours.