Headline

Building Agentic RAG @ Amgen || Lead Machine Learning Engineer || Master’s in Data Science || +91 996 385 0007

About

As a Data Scientist I apply my skills in Python and data science to develop and improve solutions that are secure, scalable, and user-friendly. I graduated with a master's degree in Data and Knowledge Engineering from Otto-von-Guericke University Magdeburg in Dec 2021, where I wrote my thesis on a novel method to detect liver fibrosis using heartbeat as excitation mechanism. I have over 6 years of work experience in data science, data analytics, and research, working with various domains, such as telecommunications, environmental science, Automotive, Automation and Pharma. I am passionate about finding innovative ways to leverage data and knowledge to solve real-world problems and create value.

Location

Hyderabad

AI stack

claude code

openai

ChatGPT

Work experience

Machine Learning Engineer at Amgen, Dec 2024 - Present · 1 yr 9 mos, Hyderabad, Telangana, India, Pipeline GPT: An LLM based chat that enables querying organization’s internal data directly within OpenAI’s ChatGPT • Designed and built an end-to-end Retrieval-Augmented Generation (RAG) pipeline, ingesting data from external APIs, chunking, transformation, and generating synthetic data to be indexed into the Databricks vector store. • Developed scalable vector ingestion jobs using PySpark and implemented Databricks Mosaic AI Vector Search to enable enterprise-grade semantic search and retrieval. • Created prompt-tuned instances of CustomGPT and integrated with Databricks vector store endpoints. • Collaborated with OpenAI and Databricks support teams to troubleshoot and resolve enterprise integration challenges, including domain authorization mismatches and performance optimization. • Led the transition from traditional Databricks vector-store-based RAG to Model Context Protocol (MCP)- based RAG to address token limitations, improve accuracy and reduce hallucination. • Architected MCP tools with Pydantic models and output parsers to guarantee consistency across all the MCP tools domain while mentoring other developers on design and implementation. • Developed MCP TestBench, an internal GenAI evaluation tool developed using DeepEval to evaluate the application performance. • Integrated MCP TestBench into GitLab’s CICD pipeline to evaluate merge requests before merging any code into the application repository. • Supported deployment teams during rollout, ensuring smooth releases, secure connectivity between components, and improved overall system reliability. • Improved deployment stability by debugging MCP client - server communication issues, optimizing Kubernetes pod allocation, and hardening MCP client configurations against network interruptions. • Implemented a multi Agent Text2Sql that can dynamically translate natural language query to SQL, validate, execute and analyze the result using Langgraph.

Data Scientist at Fortuna Identity, Feb 2022 - Dec 2024 · 2 yrs 11 mos, Hyderabad, Telangana, India, EzyBot - Visual Automation Tool • Fine-tuned YOLOv8 Image Segmentation on UI elements dataset to detect UI elements on the screen for EZYBot. • Worked on image processing algorithms using OpenCV and Google OCR, resulting in a surface automation tool named EZYBot that reduced automation script development time by 80% and eliminated coding completely. EzyBot LLM Assistant • Developing an LLM Framework that can breakdown given tasks into subtasks and resolve them by making accurate API calls from given EzyBot OpenAPI specification using LangChain. • The framework consisted of multiple agents working together like API Planner, API argument parser, API Caller and API response parser. • Developed LLM functionality to automatically fill global parameters needed for making API calls while prompting the user if data is not present. • Automatically makes requests to safe APIs while gets a confirmation from the user before making requests to sensitive APIs • Added memory functionality to remember users' chat history across multiple sessions. • Implemented output parsing to verify the correctness of the received responses. • Integrated LangSmith for tracing LLM requests and performance analytics. • Added a RAG based feature using Croma vector DB to resolve queries of Bot Developers from multiple documentations, PDFs, and manuals • Upgraded the LLM assistant from LangChain to CrewAI framework to make it more robust and resilient to queries by introducing a Manager Agent and changing the process from sequential to hierarchical. Data Analysis and Machine Learning Projects • Devised ETL pipelines to preprocess datasets, ensuring data quality and compatibility. • Created insightful visualizations of graph data using Gephi, aiding interpretation of complex clusters in the graphs. • Executed projects on Hidden Markov Models (HMMs) and LSTM networks for sequence predictions. • Deployed models as endpoints via Amazon SageMaker on AWS, created RESTful microservices using Python Flask.

Computer Vision Engineer at Inka, Apr 2021 - Jan 2022 · 10 mos, Magdeburg, Saxony-Anhalt, Germany, Responsibilities: ● Spearheaded innovative research aimed at assessing the viability of utilizing heartbeat-induced stimuli for staging liver fibrosis disease. ● Implemented the objectives of research using U-Net Encoder-Decoder (Deep Neural Networks) model for Image Segmentation and OpenCV optical flow methods for image processing in Python. ● Validated the hypothesis of the research and authored Master’s Thesis as an integral component of my academic journey.

Data Analyst at T-Systems International GmbH, Oct 2020 - Mar 2021 · 6 mos, Berlin, Germany, Responsibilities: ● Devised a use case aimed at enhancing public transport through the utilization of the International DataSpace (IDS) Connector. ● Constructed a functional web application prototype employing Python Flask, JavaScript, and OpenStreetMaps integrated with the Connector for the proposed use case. ● Implemented an advanced Named Entity Recognition (NER) model for the identification of source and destination locations within speech data, leveraging Natural Language Processing (NLP) techniques for the prototype. ● Procured relevant data by effectively communicating the significance of the use case to key representatives from over 3 mobility startups, including Tier and Voi.

Research Assistant at UFZ, Sep 2020 - Oct 2020 · 2 mos, Halle, Saxony-Anhalt, Germany, Responsibilities: ● Quantified water basin precipitation across the United States and South America through a meticulous analysis of last 75 years of historical data sourced from the MOPEX dataset. ● Developed interactive Jupyter Notebooks to effectively visualize water basins in both the US and SA, utilizing shape files. Constructed predictive models aimed at accurately estimating rainfall patterns. ● Employed these notebooks to create a comprehensive visualization of time series data, subsequently utilizing mathematical computations provided by a PhD researcher to distinguish flood and drought-prone basins.

Member Technical Staff at CDK Global, Jun 2017 - Dec 2018 · 1 y 6 mo, Hyderabad

Education

Master's degree, Data and Knowledge Engineering - Informatics/Computer Science, Otto-von-Guericke University Magdeburg

Bachelor of Technology (B.Tech.), Computer Science, Mahatma Gandhi Institute of technology

Public links

github: https://github.com/Khamar-Uz-Zama

leetcode: https://leetcode.com/UzZama