Tarun D.Private profile: Upwork shows its earnings only to signed-in members, so we check it every few weeks, not daily. MRR here is a rough estimate (Upwork's six-month earnings ÷ months active). Last read Nov 5, 2026.

Claude Certified Architect—Professional and Claude Certified Developer—Foundations. I automate complex business workflows with production AI: agentic systems, multi-agent orchestration, MCP/tool use, RAG, memory, human approvals, evaluations, guardrails, and LLMOps—not chatbot demos. I bring 11+ years across AI architecture, applied-AI product engineering, intelligent automation, and technical leadership. At Deloitte, I led enterprise GenAI and agentic-AI delivery for regulated organizations. I also founded and scaled an AI product company, and I remain hands-on from discovery and architecture through code, evaluation, deployment, and adoption. AI AUTOMATION, AGENTIC AI & APPLIED AI • Multi-agent orchestration: supervisor/worker systems, planner/executor patterns, routing, delegation, handoffs, shared state, and durable workflows • Tool-using agents: MCP, APIs, function calling, structured outputs, memory, and end-to-end business-process automation • Enterprise knowledge agents: RAG, contextual retrieval, grounded responses, citations, permissions, and human-in-the-loop review • Reliable AI systems: evaluations, guardrails, tracing, observability, retries, fallbacks, audit trails, and approval gates • Applied-AI products: Python, FastAPI, microservices, SaaS, dashboards, integrations, cloud deployment, CI/CD, and LLMOps • AI strategy and architecture: use-case discovery, model selection, cost and latency optimization, governance, and production roadmaps SELECTED AGENTIC & APPLIED-AI PROJECTS • ALQ — evaluates AI leadership through strategy simulations with unscripted agents • Cred Farmer — Claude certification preparation with 2,700+ early signups • Tiflo — specification-first agentic execution with independent verification and an auditable approval trail • NAVIK — a retrieval-grounded ship-engineering assistant using authorized manuals, alarms, and historical cases • Enterprise Banking Agentic AI — reduced manual workload by 70% • Reusable Enterprise AI Framework — adopted across engagements and reduced solution time by 60% • Agentic Workflow Builder — shortened idea-to-prototype cycles by approximately 90% • Enterprise Context Intelligence — unified context from engineering and product systems • Compliance RAG Advisor — grounded regulatory guidance in governed enterprise knowledge • Enterprise Conversational AI — 200+ voice and conversational solutions delivered for 60+ clients MEASURABLE ENGINEERING OUTCOMES • Built 50+ backend systems and established 30+ CI/CD pipelines • Reduced cloud architecture costs by 47% • Automated validation for 1,400+ Linux packages • Reduced a deployment process from eight hours to two • Won two Amazon Alexa conversational-game hackathons • Completed two Upwork projects, both with 5-star ratings CORE STACK Claude, GPT, Gemini, Mistral, LangGraph, MCP, Azure AI Foundry, Azure OpenAI, Amazon Bedrock, Vertex AI, Python, FastAPI, Django, Flask, PostgreSQL, React, Docker, Jenkins, AWS CodePipeline, and CI/CD. If you need a Claude-certified architect who can turn an ambiguous AI opportunity into a reliable, measurable production system, send me your workflow, constraints, and success metric. I can define the right automation and agent architecture—and ship it.