SUDAMA SHARMA Email : 1krishna.sudama@gmail.

com
Data Scientist (Gen AI Consultant) Github
Mobile: +91 - 8386040262 LinkedIn
Dedicated full-stack generative AI developer with 4+ years of experience in designing scalable platforms and managing
AWS deployments. Currently pursuing MTech in AI at IIT Jodhpur. Skilled in performance optimization, adapting to
dynamic environments with strong analytical abilities to enhance software reliability.
education
Indian Institute of Technology, Jodhpur Jodhpur, India
•
PG Diploma and MTech - Artificial Intelligence July 2024 - Ongoing
Jaipur Engineering College, Jaipur Jaipur, India
•
Bachelor of Technology - Computer Science June 2019
technical skills
• Languages: Python, SQL, JavaScript, TypeScript
• Skills: Data Science, AI, Machine Learning, Deep Learning, NLP, Generative AI, GPT, LLM’s, Transformer, RAG
• Frameworks: Pytorch, TensorFlow, Langchain, Langfuse, Llama Index, PySpark, Pandas, Scikitlearn, OpenCV
• Analytics: LLM FineTuning, Reranking, Predictive Modelling, Statistical Analysis, Embedding, EDA, NER, POS
• Web Dev: FastAPI, Django, Django Rest Framework, ReactJs, NodeJs, Gradio, Streamlit, Matplotlib, Seaborn, Plotly
• Tools: GIT, Bitbucket, Jira, AWS (S3, EC2, Lambda, SageMaker, Bedrock, RDS, Arora, DynamoDB, Glue)
• Practices: Object Oriented Programming, Functional Programming, Microservices, Scrum, SDLC/Agile, CI/CD.
work experience
SRM Technologies Private Limited Chennai (Remote), India
•
Data Scientist (Gen AI Consultant) July 2024 - Ongoing
◦ NL2SQL ChatBot: Developed an NL2SQL chatbot for a US e-commerce client, leveraging advanced RAG techniques,
Vanna AI, and OpenAI models. Optimized Chroma DB indexing with DDL, schema, QnSQL, and few-shot training for
PO, line items, and RMA. Implemented advanced RAG with reranker, hallucination and answer grader. Hybrid retrieval
combining vector similarity and BM25 with a reranker, enhancing accuracy through on various search methods.
◦ An in-domain classifier model was developed to filter domain-specific NL2SQL queries, blocking out-of-domain questions
to improve user experience. Metrics were also designed for QuietBot and traces generated for Snowflake integration.
◦ Skills Used: Django, Django Rest Framework, RAG, VectorDB, OpenAI, LLM, LangFuse, Langchain, Scrum, CI/CD.
Tessolve Semiconductor Pvt. Ltd. Bangalore, India
•
Software Engineer 2 (Data Scientist) Feb 2024 - July 2024
◦ Local GPT: Develop a generative AI project that allows users to query documents using LLMs even offline, adhering to
extending OpenAI API standard streaming responses. High-level API simplifies RAG pipeline by managing document
ingestion, facilitating chat and completions with contextual information. Low-level API provides advanced users flexibility
to create custom pipelines, including text-based embeddings generation and contextual chunk retrieval for queries.
◦ A functional Gradio UI client is provided for API testing, along with tools like a bulk model download script, ingestion
script, and document folder watch utility. This system enhances log analysis, supports interactive querying, and offers
intuitive visualisations, improving decision-making processes based on aggregated log data trends.
◦ Skills Used: FastAPI, Gradio UI, GenAI, LLM, Vector DB, LlamaIndex, AWS Cloud Services, Scrum.
Heptagon Technologies Pvt Ltd Bangalore, India
•
Data Scientist Nov 2022 - July 2023
◦ NPS Analysis - RAG Agent: Developed an NPS analysis system using LangChain, Chromadb, OpenAI embedding,
and LLM models. The system ingests MySQL data for aspect-based learning, offering APIs for document ingestion,
contextual completions, custom pipelines, and a Streamlit UI for testing and bulk operations.
◦ Predictive Modelling: Created predictive models using statistical analysis and ML techniques to improve customer
retention. Built data pipelines for a major Indian Media Group, predicted unknown data points with custom Rolled avg
competitor data, and optimized performance by feature engineering,hyperparameter tuning, and metrics analysis.
Cognizant Technology Solutions India Private Ltd Bangalore, India
•
Data Scientist Aug 2021 - Oct 2022
◦ Predictive Modelling: Collaborated with teams to translate requirements into actionable solutions. Performed data
exploration, feature engineering, cross-validation, and hyperparameter optimization to enhance model performance and
identify critical business variables. Developed predictive models for customer churn classification and probability scoring
using survey and metrics data. Conducted model performance evaluation and improvement analysis by leveraging AWS
SageMaker, Glue and other services for optimization.
Intelligent Retail Pvt Ltd Bangalore, India
•
Python Developer - AI/ML May 2020 - July 2021
◦ Retail Research and Analytics: Led the development of dashboard design and data collection pipeline for Ripplr’s
distribution business. Spearheaded the creation of the backend infrastructure for RIPPLR’s Champion Distribution
Management System, utilising a tech stack that included Python, MySQL, Django, React, Tableau, Seaborn, Matplotlib.
certificates
• IBM: Data Analysis Using Python
• IBM: Machine Learning with Python - Level 1
• IBM: Deep Learning using TensorFlow
• Tableau: Tableau Desktop Certified Professional