Kinshuk Dutta

AI Systems • Agentic AI • Autonomous IT • Enterprise Platforms

Building real-world AI systems where orchestration, governance, and execution actually work — beyond demos, beyond hype.

---

👋 About Me

I work at the intersection of AI systems engineering, agentic architectures, and enterprise execution models.

My focus is not just what AI can do — but how AI operates safely, scalably, and responsibly inside real organizations.

I spend my time designing and enabling:

• Agentic AI systems embedded in workflows
• Autonomous IT and AI-driven execution platforms
• Governance-first AI architectures
• Enterprise-scale AI adoption patterns

I care deeply about systems thinking, not toy models.

---

📘 Publications & Thought Leadership

• Author — AI Agents in Action
• Author — Data for AI
• Contributor to enterprise AI, data platforms, and governance discussions

My writing focuses on:

• AI operating models
• Agent orchestration and control loops
• Data integrity and AI readiness
• Execution gaps between strategy and production AI

---

🧠 What I Build

I work on and publish projects around:

• 🤖 Agentic AI architectures
• 🧩 AI orchestration & coordination patterns
• 🛡️ AI governance, safety, and control frameworks
• 🏗️ Enterprise AI system design
• ⚙️ Autonomous IT & operational intelligence

This GitHub is intentionally curated — fewer projects, higher signal.

---

⭐ Featured Repositories

These projects represent applied AI systems — designed with real-world constraints, governance, and impact in mind.

• AOS - Agent Orchestration Studio
  Agent Studio is an open-source visual playground to design, run (sandbox or BYOK), trace, and export AI agents — built to learn agentic AI by building real systems. 🔗 https://github.com/KinshukON/agent-orchestration-studio

• RAGOS - RAG Orchestration Studio
  RAG Orchestration Studio: A Control Plane for Enterprise Retrieval Architectures 🔗 https://github.com/KinshukON/rag-orchestration-studio

• Patent Navigator
  AI-powered system for navigating and analyzing patent intelligence with traceability and governance at its core.
  🔗 https://github.com/KinshukON/patent-navigator

• AI Guardian
  Safety-first AI framework to protect users and children from harmful, biased, or inappropriate AI-generated content.
  🔗 https://github.com/KinshukON/AI-Guardian

• EmotiSense
  Hybrid neuro-symbolic emotion detection system focused on explainability, personalization, and responsible AI.
  🔗 https://github.com/KinshukON/emotisense

• AgriRescueNet
  AI-driven disaster response and resilience platform for agriculture and food systems.
  🔗 https://github.com/KinshukON/AgriRescueNet

• FairQuote
  Fairness-aware automated pricing and quotation system designed to reduce bias and improve transparency.
  🔗 https://github.com/KinshukON/fairquote

(Repos will evolve deliberately — quality over quantity.)

---

🧭 Philosophy

• AI should augment execution, not create chaos
• Governance must be embedded, not bolted on
• Agents must be observable, accountable, and interruptible
• Systems matter more than models

If it can’t survive contact with production — it doesn’t ship.

---

🔗 Connect

• 🌐 Blog: https://datanizant.com
• 💼 LinkedIn: https://www.linkedin.com/in/kinshuk-dutta
• 📚 Books & Writing: linked from blog

---

This profile is a living system — it will evolve as the work does.