V Priyanka

AI/ML Engineer with 4 years of experience building GenAI and data science solutions in financial services and healthcare. Specialized in agentic AI (CrewAI, LlamaIndex), RAG pipelines, and LLM-powered NLP pipelines (text summarisation, question answering, entity extraction), with hands-on experience taking projects from PoC to production. Proven track record in client-facing delivery, agentic system development, cloud-based ML pipelines (GCP, Vertex AI), and responsible AI practices in regulated healthcare environments.

Strong expertise in Python, modern GenAI stacks, and end-to-end system design

Software Developer (AI) at Prodapt (2025-03 – Present)

• Led a 3-member team in transforming a proof-of-concept data migration solution into a production-ready, platform-agnostic agentic framework. Designed multi-agent workflows with human-in-the-loop validation to automate source-to-target table migration, data transformation, and quality checks, scaling the solution from a limited PoC to support enterprise-scale migrations.
• Built LlamaIndex-based agentic workflows for automated code modernization, including PySpark and Scala optimization and code transformation. Introduced a human-in-the-loop validation stage for reviewing AI-generated design specifications before generating target implementations, improving reliability and reducing migration errors.
• Built an LLM-powered document generation pipeline ingesting complex nested JSON; implemented structured outputs, prompt orchestration, and hallucination checks via dual-layer clinical review — validating medication names, diagnoses, and patient events across hundreds of cases with responsible AI practices in a regulated healthcare environment.
• Containerized a multi-service GenAI application exposing REST APIs via FastAPI and Flask,configured inter-service networking in Docker, implemented request validation, exception handling, and logging delivering client-ready images for self-hosted production deployment.

Associate Data Scientist at Techvantage.ai (2022-08 – 2025-03)

• Experience in using Agentic Frameworks like CrewAI and created multiple agents for automating policy renewal and churn prediction by empowering agents to work together seamlessly, tackling complex tasks.
• Built GenAI applications across healthcare and insurance using open-source LLMs (LLaMA 2, Falcon) and enterprise models (GPT-4, MedLM, Gemini); use cases included document Q&A, summarisation, classification, and data extraction — implemented via LangChain, Haystack, and FAISS-backed semantic search.
• Implemented context-aware chatbot interactions for medical research, insurance claims, and legal document analysis, utilizing LLMs with memory storage via LangChain and MongoDB to enable multi-turn conversations in document-based Q&A systems.
• Optimized cloud-based AI models and pipelines using GCP, Apache Beam, and scalable ETL workflows for efficient data processing, automation, and deployment, while developing machine learning models for risk prediction and recommendation systems, including a driver risk assessment model with supervised learning (RandomForest) and evaluated using precision, recall, and F1 metrics.

B.Tech in Electronics and Communication – Model Engineering College (2016 – 2020)

M.Tech in AI and Data Science – Muthoot Institute of Science and Technology (2020 – 2022)