Rohan Sitaniya

About Rohan Sitaniya

AI Engineer building enterprise-scale GenAI systems at the intersection of Agentic AI, workflow automation, and intelligent infrastructure. Over the last 5+ years, I’ve worked across finance, healthcare, and enterprise SaaS domains designing production-grade AI systems spanning multi-agent orchestration, RAG pipelines, AI governance, fraud intelligence, underwriting automation, and API-native AI platforms. Currently at Jitterbit, I build AI-powered iPaaS and APIM systems enabling enterprises to create workflows, integrations, and APIs through natural language using multi-agent architectures, function calling, hybrid retrieval, and scalable orchestration frameworks. Previously at American Express, I worked on GenAI governance, financial risk intelligence, fraud detection, and enterprise NLP systems — building Responsible AI frameworks focused on hallucination detection, observability, auditability, compliance, and production AI safety for regulated environments. I’m particularly interested in the future of enterprise AI agents, autonomous business workflows, and scalable AI infrastructure that can reliably operate in real-world production systems.

Experience

1. Ai Engineer

   Jitterbit

   Oct 2024 — Present

   APIM Bot: Built a conversational assistant for Jitterbit’s API Manager platform, automating complete API lifecycle management via natural language interaction. Implemented token-aware summarization & memory reducing LLM cost by ~60 %, handling 10 K+ daily queries with low-latency streaming. Developed provider-agnostic LLM adapter (OpenAI ↔ Llama ↔ Bedrock) reducing init overhead ~80 %. Ensured Responsible AI through encrypted auth, bias & hallucination filters, Langfuse tracing, zero LLM data exposure.iPaaS Bot: Built multi-agent architecture for integration automation — Orchestrator, NLU (Intent & Entity), Planner (RAG Agent), Executor. The Planner used ChromaDB semantic search to map workflow requirements from existing integration patterns, improving context precision ~50 % and cutting retrieval latency ~40 %. Executor handled 50+ node graphs with checkpointing & resumability. Delivered 99.5 % uptime via Kubernetes microservices and CI/CD pipelines.

Education

• Indian Institute of Technology, Kharagpur

  Integrated Dual Degree (B.Tech + M.Tech), Biotechnology and Biochemical Engineeering

  2015 — 2020

Skills

• Leadership
• Microsoft Excel
• Html
• Microsoft Office
• Management
• Microsoft Word
• Adobe Photoshop
• Teamwork
• Public Speaking
• Microsoft Powerpoint
• C
• Programming