Abhishek Mamdapure
Abhishek Mamdapure
Professional Journey
- Spearheaded the development and deployment of a Generative AI project (Google Gemini based) within a large-scale customer-facing program (IVR).
- Developed proprietary solutions tailored to serve a vast customer base across multiple geographies.
- Built the end-to-end AI pipeline, including backend services using FastAPI, ensuring scalable, production-ready architecture.
- Built and mentored a high-performing team to execute the project vision and deliver under tight timelines.
- Delivered multi-million dollar cost savings through automation and AI-driven efficiencies in customer service workflows.
- Pioneered the adoption of Generative AI at Citi, establishing systems and standards for future AI initiatives.
- Drove AI/ML initiatives from prototype to production in close collaboration with leadership.
- Developed and deployed embeddings-based search engines using FAISS, Pinecone, and MongoDB Atlas in RAG architectures.
- Built NLP models for medical entity extraction from biomedical abstracts to enable downstream analytics.
- Designed and implemented a knowledge graph pipeline over massive unstructured life sciences datasets.
- Created an AI chatbot that generates contextual graphs and plots from user queries for pharma analytics use cases.
- Architected and maintained scalable data pipelines and mentored junior data scientists.
- Analyzed large datasets to identify trends, patterns, and relationships.
- Designed and implemented data pipelines to collect, store, and process large datasets.
- Developed dashboards and reports to communicate insights to stakeholders.
- Improved data analysis processes and tools as part of the core data team.
- Designed and implemented an NLP system to convert natural language questions into SQL queries for a target database.
Projects
Generative AI Customer Service (IVR)
Role: Officer - GenAI Development @ Citibank
Spearheaded a Generative AI project using Google Gemini within a large-scale customer-facing program. Built an end-to-end AI pipeline with FastAPI and delivered significant cost savings through automation.
RAG Search Engine
Role: Lead Data Scientist @ Pienomial
Developed and deployed embeddings-based search using FAISS, Pinecone, and MongoDB Atlas. Powered fast, context-aware retrieval using RAG architecture.
Biomedical NLP and Entity Extraction
Role: Lead Data Scientist @ Pienomial
Built NLP models to extract medical entities from biomedical literature, enabling search and analytics for healthcare and life sciences.
Knowledge Graph Pipeline
Role: Lead Data Scientist @ Pienomial
Designed a knowledge graph pipeline over large unstructured datasets to represent complex biomedical relationships for advanced discovery.
AI Chatbot for Pharma Analytics
Role: Lead Data Scientist @ Pienomial
Created a chatbot that generates contextual graphs and plots from user queries to support pharma sales reps with quick insights.
Malicious URLs Detection
Academic Project (M.Tech)
Machine learning based system to detect malicious URLs by analyzing URL features and corresponding website characteristics.
Crowd Size Estimator with PyTorch
Personal Project
Crowd counting using CERNet (CNN with dilated convolutions). Implemented in PyTorch and deployed via Flask.
TLDR Text Summarizer
Personal Project
Text summarization based on sentence similarity ranking, implemented with NLTK and deployed using Flask.
Education and Certifications
Education
Subjects: Machine Learning, Operations Research, Optimization, Numerical Computing.
Skills
Certifications and Papers
- AI Engine for research and modelling in field of Immunotherapy
- NPTEL - Deep Learning, Indian Institute of Technology Kharagpur
- NPTEL - Practical Machine Learning With Tensorflow, Indian Institute of Technology Madras
- Data Science Math Skills, Duke University via Coursera
- Statistical Learning, Stanford University
- Python for Data Science and Machine Learning, Udemy