Summary
Overview
Work History
Education
Skills
Certification
Languages
Publications Research
Timeline
Generic

Bhargav Solanki

Winterthur

Summary

Experienced data scientist specializing in the deployment and optimization of large language models (LLMs), generative AI solutions, and advanced RAG systems. Proficient in cloud infrastructure, scalable machine learning (ML) solutions, and improving model performance to generate impactful data-driven insights. Recognized for driving innovation and facilitating digital transformation within enterprise applications. Published researcher with extensive knowledge in extreme multi-label classification and advanced natural language processing (NLP) techniques.

Overview

10
10
years of professional experience
1
1
Certification

Work History

Data Scientist

Abacus Research AG
07.2023 - Current
  • LLM Optimization: Fine-tuned advanced LLMs (e.g., Mistral, LLaMA, Qwen) using LoRA and QLoRA, achieving significant accuracy improvements from 70% to 90%
  • RAG System Development: Built sophisticated Retrieval-Augmented Generation (RAG) systems using LangChain, implementing smart chunking and GraphRAG for improved information retrieval and relevance in NLP applications
  • Model Deployment: Deployed fine-tuned models with vLLM, LMDeploy, leveraging Python, PyTorch, and Hugging Face for scalable, OpenAI-compatible deployments
  • Information Extraction: Developed multimodal and text-only models to extract structured data from documents, optimizing processing efficiency
  • Experimental ML Deployment: Conducted extensive model experiments on Hugging Face and ModelScope, deploying models with Gradio for versatile applications
  • Image Processing: Led projects on image classification and object detection using Detectron and YOLO, enhancing Abacus' visual data capabilities
  • Data Management: Spearheaded data cleaning, preprocessing, and feature engineering, ensuring quality insights from diverse data sources

Software Engineer 2

Appian Corporation
11.2020 - 09.2021
  • Developed runtime code and security tools for SAIL developers, improving code quality and minimizing debugging time saving 5 seconds of load time per interaction
  • Created automation features for application object generation, streamlining development workflows
  • Prototyped SAIL capabilities on AWS using Lambda and DynamoDB, optimizing resources with serverless architecture

Technical Lead, Software Development Engineer

1Ticket.com
12.2017 - 11.2019
  • Team Leadership: Led a team of 4 engineers, overseeing project timelines, task delegation, and mentoring to ensure high-performance and on-time delivery of technical solutions
  • Microservices Development: Architected and developed microservices on AWS, integrating ticket broker APIs through REST services using AWS CloudFormation, CloudWatch, and DynamoDB
  • System Integration: Designed and implemented adapter services in Python, facilitating seamless system integration with Microsoft SQL and enhancing cross-functional communication with external marketplaces

Software Development Engineer In Test

comScore, Inc.
04.2017 - 12.2017
  • Enhanced iOS/Android app features using Swift, Objective-C, and Python; maintained database applications, improving efficiency with automation scripts

Data Scientist Intern

Intuit
07.2014 - 12.2014
  • Predictive Modeling: Developed predictive models for business applications, leveraging data-driven insights to support key business objectives
  • Command-Line Tool Development: Built a command-line tool for streamlined model deployment, facilitating smoother transitions from development to production environments
  • Research & Development: Conducted research on advanced machine learning techniques, contributing to the team's data science capabilities in predictive analytics

Education

Master of Science - Data Analytics and Economics

BeNeFri Consortium, University of Fribourg
Fribourg, Switzerland
06.2024

Master of Science - Computer Science and Big Data Analytics

Rochester Institute of Technology
Rochester, New York
01.2016

Bachelor of Engineering - Information Technology

K.J. Somaiya College of Engineering
Mumbai, India
01.2013

Skills

  • Machine Learning & AI, Generative AI, LLMs, Text, Image, MultiModal, LLM fine-tuning, LoRA, RAG,
  • AWS, Lambda, CloudFormation, Sagemaker, Python, Pandas
  • Hugging Face, Gradio, VLLM, PyTorch, LangChain, Linear algebra
  • Data cleaning, Preprocessing, Feature engineering, Deep Learning, Complex Problem-Solving
  • Agile Methodologies, Data science research methods
  • Continuous deployment, Team building

Certification

  • AWS Developer Associate
  • Microeconomics of Competitiveness, Harvard Business School

Languages

English
Bilingual or Proficient (C2)
German
Elementary (A2)

Publications Research

  • AttentionXML vs. LLMs: An Empirical Evaluation of Extreme Multi-Label Classification Techniques, Big Data Conference 2024, Washington DC, forthcoming publication, Investigates classification techniques for high-dimensional datasets, providing insights into LLM vs. traditional model capabilities.
  • Response and Recovery: A Quantitative Approach to Emergency Management, Developed a quantitative framework to aid disaster response agencies in efficient decision-making during emergencies.

Timeline

Data Scientist

Abacus Research AG
07.2023 - Current

Software Engineer 2

Appian Corporation
11.2020 - 09.2021

Technical Lead, Software Development Engineer

1Ticket.com
12.2017 - 11.2019

Software Development Engineer In Test

comScore, Inc.
04.2017 - 12.2017

Data Scientist Intern

Intuit
07.2014 - 12.2014
  • AWS Developer Associate
  • Microeconomics of Competitiveness, Harvard Business School

Master of Science - Data Analytics and Economics

BeNeFri Consortium, University of Fribourg

Master of Science - Computer Science and Big Data Analytics

Rochester Institute of Technology

Bachelor of Engineering - Information Technology

K.J. Somaiya College of Engineering
Bhargav Solanki