Experience – Omer Chaudhry

Data Science Institute Brown University,

Undergraduate Researcher

Employed machine learning frameworks, including PyTorch, Scikit-Learn, and JAX, to develop and enhance predictive
models, leveraging GPU-accelerated optimization and performance tracking via Weights & Biases.
Developing scGrapHiC, a deep learning framework integrating Graph Attention Networks, transformer encoders, and
positional encodings to predict pseudo-bulk single-cell Hi-C (scHi-C) contact maps from scRNA-seq data increasing
accuracy by 15%
Exploring LLM-inspired encoders for feature extraction and representation learning in genome-wide structural prediction.

Conversational AI Lab Brown University,

Developer

Engineered Smart GenAI systems, leveraging AutoGen for agent coordination and SQLite for data persistence, optimizing
real time workflows through autonomous AI-driven decision-making.
Designed and implemented LangChain-powered Agents, integrating LLM-based keyword extraction, frequency computation,
and prompt chaining to autonomously generate structured compositions using GPT-4o-mini.
Conducted experiments with parameter-efficient fine-tuning techniques (e.g., LoRA, QLoRA) on open-weight LLMs,
benchmarking performance using BLEU, ROUGE, and perplexity scores to optimize task-specific text generation.

Infinit Labs

Technical Project Management Intern

Contributed to developing the first Urdu and Punjabi Language chatbots, integrating IBM Watson APIs and NLP models for multilingual communication. Designed and implemented API endpoints and authentication layers in Python and Flask, optimizing response times by 35%. Developed unit and functional test suites using PyTest and JUnit, improving software reliability.

Digital DirAction,

Lead Developer & Co-founder

Designed and developed AI-driven applications to improve workflow automation for businesses, leading to 11% cost
reduction for clients annually.
Built an LLM-powered RAG system for enterprise knowledge management using OpenAI embeddings and Llama3,
optimizing retrieving internal information and generating standardized communications.
Integrated machine learning models (Logistic Regression, Random Forest, Gradient Boosting) into predictive analytics
pipelines, achieving 85% accuracy for medical diagnostics.

IBM Data Science Lab,

Professional AI Partner

Created data-driven storytelling dashboards with Matplotlib, Seaborn, and Folium, transforming raw datasets into actionable
insights and automating reports with Python scripting
Developed foundation models and LLMs, leveraging Watsonx.ai and OpenAI APIs to develop chatbots, AI-powered
assistants, and generative AI applications using Python, Flask, and RESTful APIs.
Implemented end-to-end data processing workflows using NumPy, and SciPy, optimizing data wrangling, feature
engineering, and exploratory data analysis (EDA) for real-world business problems.

Neo News Network,

TV Show Host -> Tech Strategy Advisor

Recorded over 300 episodes of the talk show "Halwa Puri" for Neo News, a national news channel as Asia’s Youngest TV
Show Host for 5 years - interviewing prominent figures, including the Prime Minister and other notable personalities.