Kunj Patel
AI Engineer building GenAI, Agentic & LLMOps systems on GCP/AWS. Deployed production AI sales agents & legal RAG pipelines at Vypar TaxOne using Python, LangGraph & AWS Bedrock. IEEE-published RAG researcher (ICCES 2024).
Hire this person
Send a job offer directly to this candidate
About
AI Engineer building and shipping production-grade agents and RAG pipelines on GCP and AWS – specialising in LangChain, Vertex AI, and multi-agent orchestration. Delivered 5+ agentic systems in production environments. Certified Google Cloud ML Engineer and AWS GenAI practitioner.
Experience
AI Engineer at Vyapar TaxOne (2025-11 – Present)
Building and shipping production-grade AI agents and RAG systems
- Built an AI sales agent using LangChain Agents that processes live call transcripts, classifies buying intent (0–10), extracts key takeaways, and automatically adds structured results to a training dataset – building a corpus of 500+ labelled calls to enable continuous improvement of the sales pipeline.
- Delivered the GST Compliance Assistant, a domain-specific Retrieval-Augmented Generation (RAG) system on Vertex AI enabling Chartered Accountants to retrieve precise answers from dense legal documents – served as a scalable, serverless system on Cloud Run with reliable auto-scaling under variable load.
- Developed a self-healing extraction agent (7-stage pipeline) that ingests bank statements from any institution, auto-detects format, and autonomously rewrites failing scripts on error – retrying until output passes schema validation with zero manual intervention.
AI-Enhanced UI/UX Developer at TST Technology (2025-01 – 2025-10)
Designing intelligent UI/UX flows for AI-driven platforms
- Designed intelligent UI/UX flows for AI-driven platforms and automated image generation using N8N and LLMs, reducing manual effort by 60%; awarded Best Performer of the Quarter.
Education
B.E. in Computer Science and Design – A.D. Patel Institute of Technology (2021 – 2025)
