{
  "schemaVersion": 1,
  "identity": {
    "name": "Shamil Sherhan",
    "title": "Senior AI Systems Engineer | Agentic AI | LLM Applications | Data Engineering"
  },
  "professionalSummary": [
    "AI systems engineer with 7+ years of experience spanning data integration, cloud data warehousing, and production-grade LLM applications for a Fortune 500 automotive parts distribution client. Progressed from Informatica ETL development through Snowflake and BigQuery engineering to leading a team of five engineers building multi-agent AI systems that process 500K+ queries monthly.",
    "Reduced LLM API costs by 30–50% and improved system reliability by 40%.",
    "Strong in backend engineering (Python, FastAPI, Docker) and data infrastructure (BigQuery, Snowflake, Informatica, PySpark). Seeking senior individual contributor, technical lead, or consulting roles with ownership of AI architecture from design through production."
  ],
  "skills": [
    {
      "label": "AI/LLM Engineering",
      "skills": [
        "AI agents",
        "Anthropic Claude",
        "Azure OpenAI",
        "Chroma",
        "Context management",
        "Deterministic AI pipelines",
        "Evaluations",
        "Graph retrieval",
        "Guardrails",
        "LLM applications",
        "Memory systems",
        "Multi-agent systems",
        "OpenAI",
        "Pinecone",
        "Prompt engineering",
        "RAG",
        "Reranking",
        "Weaviate"
      ]
    },
    {
      "label": "Frameworks and Tools",
      "skills": [
        "Airflow",
        "AutoGen",
        "CI/CD",
        "CrewAI",
        "dbt",
        "Docker",
        "FastAPI",
        "Git",
        "Kubernetes",
        "LangChain",
        "LangGraph",
        "LlamaIndex",
        "Prefect",
        "Pydantic"
      ]
    },
    {
      "label": "Data Engineering",
      "skills": [
        "Advanced SQL",
        "BigQuery",
        "Data modeling",
        "Data quality validation",
        "ETL/ELT pipelines",
        "Informatica Cloud (IICS)",
        "Informatica PowerCenter",
        "PySpark",
        "Query optimization",
        "Snowflake",
        "Storage optimization"
      ]
    },
    {
      "label": "Backend and Systems",
      "skills": [
        "API design",
        "CLI tooling",
        "Distributed systems",
        "Fault tolerance",
        "Linux",
        "Observability",
        "Python",
        "REST APIs",
        "Terraform",
        "Workflow orchestration"
      ]
    },
    {
      "label": "Methodologies",
      "skills": [
        "Agile",
        "Code review",
        "Confluence",
        "Jira",
        "Scrum",
        "Team leadership",
        "Technical documentation"
      ]
    }
  ],
  "experience": [
    {
      "title": "AI Systems Engineer",
      "period": "2025–Present",
      "achievements": [
        "Lead a team of five engineers building and deploying multi-agent AI systems for enterprise workflow automation across automotive parts distribution, supply chain optimization, and inventory management.",
        "Architected and deployed 10+ multi-agent systems for structured task execution, enabling autonomous decision-making across complex, multi-step processes serving 500K+ monthly queries.",
        "Built hybrid memory systems combining short-term context with long-term vector and graph retrieval using Pinecone and Chroma, supporting 50+ concurrent user sessions with sub-two-second response times.",
        "Reduced LLM API calls by 30–50% through deterministic preprocessing pipelines that resolve unambiguous cases before model invocation, saving an estimated $15K–$25K per month.",
        "Improved AI output reliability by 40% through structured Pydantic schemas, guardrails, and multi-layer validation across 8+ production AI services.",
        "Designed reusable, modular AI architectures adopted by 4+ engineering teams, reducing onboarding and development time for new AI features by 60%.",
        "Built high-throughput Python services handling 1M+ daily requests with logging, tracing, and metrics; containerized them with Docker and deployed them on Kubernetes."
      ]
    },
    {
      "title": "Google BigQuery Engineer",
      "period": "2024–2025",
      "achievements": [
        "Designed and optimized BigQuery architectures for supply chain analytics and parts catalog management, processing 5TB+ daily with 20–40% query-performance improvements through partitioning, clustering, and materialized views.",
        "Reduced data infrastructure costs by 15–30%, saving $10K–$20K monthly through storage tiering, slot reservations, query cost controls, and removal of redundant processing.",
        "Built 30+ dbt transformation models powering inventory analytics, demand forecasting, and real-time supply chain visibility for 200+ business users.",
        "Developed automated data-quality validation across 100+ pipelines, reducing manual oversight by 50% and achieving 99.5% downstream data accuracy.",
        "Created internal CLI tooling for pipeline monitoring, alerting, and operational reporting, reducing mean issue-resolution time from 45 minutes to under 15 minutes."
      ]
    },
    {
      "title": "Snowflake Engineer",
      "period": "2021–2024",
      "achievements": [
        "Led a team of 10 engineers designing scalable Snowflake warehouse architectures for enterprise analytics across parts distribution, supply chain, and sales.",
        "Reduced Snowflake compute and storage costs by 25–35% through warehouse sizing, auto-suspend policies, resource monitors, and query profiling.",
        "Built end-to-end ELT pipelines with Snowflake, dbt, and Airflow for batch and near-real-time ingestion from 50+ ERP, CRM, and POS source systems.",
        "Developed advanced SQL models using star and snowflake schemas plus Snowflake-native streams, tasks, and dynamic tables for downstream BI and ML workflows.",
        "Reduced engineering effort by 30% through modular framework design, enabling 15+ new pipeline types without greenfield builds.",
        "Mentored junior engineers through code reviews, pair programming, technical knowledge-sharing, and Jira-based sprint workflows."
      ]
    },
    {
      "title": "Informatica Developer",
      "period": "2018–2021",
      "achievements": [
        "Designed and developed 50+ Informatica PowerCenter and IICS workflows for enterprise integration across on-premises and cloud environments supporting parts catalog and order management systems.",
        "Migrated legacy ETL processes to Informatica Cloud (IICS), reducing execution time by 40% and improving scheduling reliability.",
        "Built reusable mappings, mapplets, and session configurations that standardized ETL development and reduced development time by 35%.",
        "Implemented validation and reconciliation frameworks that maintained 99%+ source-to-target accuracy for critical sales, inventory, and distribution reporting.",
        "Managed and optimized batch workflows handling 10M+ records daily with 99.5% SLA compliance."
      ]
    }
  ],
  "projects": [
    {
      "title": "Multi-Agent Supply Chain Q&A",
      "description": "LangGraph-based multi-agent system with Router, SQL Agent, RAG Agent, and Summarizer for intelligent supply chain question answering.",
      "technologies": [
        "ChromaDB",
        "FastAPI",
        "LangGraph",
        "Python",
        "SQLite"
      ],
      "url": "https://github.com/shamilsherhan/multi-agent-supply-chain-qa"
    },
    {
      "title": "Hybrid RAG Retrieval Pipeline",
      "description": "Production-grade RAG pipeline combining vector search with knowledge graph traversal for superior retrieval accuracy and context.",
      "technologies": [
        "ChromaDB",
        "FastAPI",
        "LlamaIndex",
        "NetworkX",
        "Python"
      ],
      "url": "https://github.com/shamilsherhan/rag-hybrid-retrieval-pipeline"
    },
    {
      "title": "Snowflake dbt Analytics",
      "description": "End-to-end dbt project with star schema models, comprehensive tests, and documentation for supply chain analytics on Snowflake.",
      "technologies": [
        "dbt",
        "Python",
        "Snowflake",
        "SQL"
      ],
      "url": "https://github.com/shamilsherhan/snowflake-dbt-analytics"
    }
  ],
  "caseStudies": [
    {
      "title": "Building an Evidence-First Knowledge Assistant",
      "description": "How a basic internal search interface evolved into a conversational assistant with source authority, evidence-level citations, durable history, and bounded live retrieval.",
      "date": "2026-09-24T00:00:00.000Z",
      "tags": [
        "Knowledge Retrieval",
        "LLM Agents",
        "Observability",
        "RAG",
        "Search"
      ],
      "url": "https://sherhan.me/case-studies/evidence-first-knowledge-assistant/"
    },
    {
      "title": "Generating UI Prototypes from a Living Design System",
      "description": "A UI-first workflow that snapshots the real application, converts screens into compact semantic templates, and lets an LLM make bounded changes instead of inventing the product from scratch.",
      "date": "2026-09-23T00:00:00.000Z",
      "tags": [
        "Automation",
        "Design Systems",
        "LLM Agents",
        "Playwright",
        "UI Prototyping"
      ],
      "url": "https://sherhan.me/case-studies/living-design-system-prototypes/"
    },
    {
      "title": "Designing Safe MCP Servers for Enterprise Tools",
      "description": "How I exposed issue tracking, documentation, and messaging systems to coding agents without turning natural-language instructions into unrestricted write access.",
      "date": "2026-09-22T00:00:00.000Z",
      "tags": [
        "Automation",
        "Model Context Protocol",
        "Python",
        "Security",
        "Tool Use"
      ],
      "url": "https://sherhan.me/case-studies/safe-enterprise-mcp-servers/"
    },
    {
      "title": "From Git Push to Approved Worklog",
      "description": "A local-first assistant that observes pushed commits, matches them to active work, drafts progress updates, and waits for explicit human approval before posting anything.",
      "date": "2026-09-21T00:00:00.000Z",
      "tags": [
        "Automation",
        "Developer Tools",
        "Git",
        "Go",
        "Human-in-the-Loop"
      ],
      "url": "https://sherhan.me/case-studies/human-in-the-loop-commit-assistant/"
    },
    {
      "title": "A Conversational Ticket Assistant: From Plans to Tool Chaining",
      "description": "A chat assistant for a ticket system that started with explicit approvable plans and evolved into a bounded tool loop with deterministic shortcuts, independent criticism, and UI-first prototyping.",
      "date": "2026-09-20T00:00:00.000Z",
      "tags": [
        "Conversational AI",
        "Design Automation",
        "Human-in-the-Loop",
        "LLM Agents",
        "Tool Use"
      ],
      "url": "https://sherhan.me/case-studies/conversational-ticket-assistant/"
    },
    {
      "title": "Agentic SDLC Platform: From Ticket to PR",
      "description": "A multi-agent platform that carries a ticket from intent to merged PR: conversational story authoring, planning, code execution in the right repo, and a separate critic quality gate, with a human approving plans and handling escalations at the moments that matter.",
      "date": "2026-06-02T00:00:00.000Z",
      "tags": [
        "Automation",
        "DevTools",
        "FastAPI",
        "LLM Orchestration",
        "Multi-Agent Systems",
        "Quality Gates"
      ],
      "url": "https://sherhan.me/case-studies/agentic-sdlc-platform/"
    },
    {
      "title": "Retail Analytics for an Imported Candy Store",
      "description": "Built inventory, demand forecasting, and basket analysis systems for an imported candy store to guide purchasing, anticipate short-term demand, and improve cross-sell decisions.",
      "date": "2026-04-26T00:00:00.000Z",
      "tags": [
        "Basket Analysis",
        "Data Engineering",
        "Decision Systems",
        "Demand Forecasting",
        "Inventory Optimization",
        "Retail Analytics"
      ],
      "url": "https://sherhan.me/case-studies/retail-analytics-imported-candy-store/"
    },
    {
      "title": "Agentic RAG with Iterative Refinement",
      "description": "A hybrid retrieval service for code, tickets, and docs — with an LLM that critiques its own results, auto-fetches parent classes or config files when a chunk is incomplete, and refines the query when results miss.",
      "date": "2026-04-18T00:00:00.000Z",
      "tags": [
        "Hybrid Search",
        "Knowledge Retrieval",
        "LLM Agents",
        "pgvector",
        "RAG"
      ],
      "url": "https://sherhan.me/case-studies/agentic-rag-retrieval/"
    },
    {
      "title": "Automated Data Pipeline with Quality Monitoring",
      "description": "How I replaced a manual weekly reporting workflow with an automated ETL pipeline and live dashboards, then added a data quality monitoring layer that validates and alerts on the data flowing through it.",
      "date": "2026-01-14T00:00:00.000Z",
      "tags": [
        "Analytics",
        "Automation",
        "Data Engineering",
        "Data Quality",
        "ETL",
        "Monitoring"
      ],
      "url": "https://sherhan.me/case-studies/automated-data-pipeline/"
    },
    {
      "title": "From Minutes to Seconds: Optimizing SQL & BigQuery Performance at Scale",
      "description": "How I cut query time and cloud costs by fixing inefficient SQL and improving data warehouse performance.",
      "date": "2025-09-01T00:00:00.000Z",
      "tags": [
        "BigQuery",
        "Data Engineering",
        "Optimization",
        "SQL"
      ],
      "url": "https://sherhan.me/case-studies/sql-bigquery-optimization/"
    }
  ],
  "posts": [
    {
      "title": "Testing Non-Deterministic Agents in CI",
      "description": "A practical design for combining deterministic contracts, semantic checks, DeepEval, Promptfoo, and dual LLM judges without letting flaky evaluation hold every pull request hostage.",
      "date": "2026-09-19T00:00:00.000Z",
      "tags": [
        "CI-CD",
        "DeepEval",
        "LLM Evaluation",
        "Promptfoo",
        "Testing"
      ],
      "url": "https://sherhan.me/blog/testing-nondeterministic-agents-in-ci/"
    }
  ],
  "elsewhere": [],
  "contact": [
    {
      "label": "Website",
      "url": "https://sherhan.me/"
    },
    {
      "label": "Email",
      "url": "mailto:shamil.sherhan@gmail.com"
    },
    {
      "label": "LinkedIn",
      "url": "https://linkedin.com/in/shamilsherhan"
    }
  ]
}