Hi, I'm Munazzah Rakhangi

AI Engineer

Building LLM-powered and agentic AI systems, full-stack platforms, and production ML pipelines for legal tech, healthcare, and e‑commerce.

About Me

I’ve spent the last couple of years building ML systems, LLM-powered apps, and data pipelines—and I still get a kick out of seeing something go from idea to production.

These days I’m deep in agentic AI and full-stack: Python, PyTorch, FastAPI, React. I’ve led small teams and shipped AI in legal tech, healthcare, and e‑commerce. What I care about most is building things that are reliable, useful, and actually solve a problem—not just technically impressive.

Machine Learning & AI

I love turning messy data into models that actually work—whether it’s prediction, NLP, or computer vision. PyTorch and scikit-learn are my go-tos, and I care about shipping things that hold up in production.

Agentic AI & LLMs

RAG, multi-agent workflows, and LLM orchestration are where I spend a lot of my time now. I’ve been building agentic systems for legal intake and automation—tool-calling, structured outputs, and workflows that run on their own.

Full Stack & Data

From React and FastAPI to PostgreSQL and AWS—I’m comfortable across the stack. I like owning the whole pipeline: ETL, APIs, and deployment, so the thing actually runs and stays up.

Leading & Shipping

I’ve led small engineering teams and shipped AI systems in legal tech, healthcare, and e‑commerce. For me it’s about clear communication, reliability, and making sure what we build actually helps people.

Experience

My professional journey and key accomplishments

AI/ML Engineer

ReferU.AI

February 2026 - Present
Remote USA
  • Developing agentic AI legal intake system using LLM workflows and tool-calling to automate client screening, data extraction, and structured case intake
  • Architecting multi-agent pipelines for document analysis, case classification, and attorney matching using Python, FastAPI, and LLM orchestration
  • Building production AI services integrating LLM APIs and structured outputs to automate legal workflows and reduce manual case processing

Software Engineer Intern

Prox Shopping

September 2025 - Present
Los Angeles, CA, USA (Remote)
  • Architected NLP pipeline using OpenAI APIs to convert 50+ weekly retailer flyers into structured JSON datasets with 95%extraction accuracy.
  • Built scalable Python web scraping and distributed task system automating flyer ingestion and reducing manual data collection by 80%.
  • Designed ETL pipelines and PostgreSQL/Supabase backend enabling centralized data storage and automated email deliverywith 99.5% uptime

Team Lead and Full stack developer

Open Source with SLU

Aug 2025 - December 2025
Saint Louis, MO, USA
  • Architect end-to-end system design across frontend, backend, API, and ML modules
  • Engineer React.js frontend with Tailwind CSS and multi-language support using react-i18next
  • Design FastAPI/Node.js backend with PostgreSQL and JWT authentication for secure user management
  • Direct API integration and establish GitHub workflows (branching, PRs, code reviews, CI)
  • Collaborate on ML modules (Heart Risk Prediction, Prescription Detection) for predictive analytics integration

Data Analyst

Excelerate Data Visualization Internship

February 2025 – April 2025
Remote, USA
  • Analyzed Facebook ad campaigns using Google Sheets, Power BI, and Excel to uncover engagement trends
  • Optimized budget allocation by evaluating campaign performance, improving efficiency by 15%
  • Created interactive dashboards and reports for strategic decision-making using Power BI and Excel

Application Engineer

Newgen Software Technologies

July 2021 – December 2022
India
  • Enhanced SmartGBO transaction software for ICICI Bank, improving user experience using Java, JavaScript, and Apache Tomcat
  • Collaborated on design and testing to boost system efficiency and reduce errors with Oracle Database and SQL
  • Conducted root-cause analysis, reducing recurring defects by 20% through automated testing and debugging
  • Debugged code and analyzed logs to identify and fix performance bottlenecks using JIRA and Git

Skills

My expertise and technical proficiencies

Languages & Runtimes

Python
SQL
JavaScript
TypeScript

Frontend

React.js
Tailwind CSS
HTML/CSS

Backend & APIs

FastAPI
Node.js
Flask
Express.js
REST APIs

Agentic AI & LLMs

OpenAI APIs
LLM Orchestration
RAG
LangChain
Prompt Engineering
Schema-Constrained Generation

ML & Data

PyTorch
TensorFlow
Scikit-learn
Pandas
NumPy

Databases

PostgreSQL
MySQL
MongoDB
Supabase

Cloud & DevOps

AWS EC2
AWS S3
Docker
Git
CI/CD

Projects

A selection of my personal and Academic projects

TaskScribe – AI meeting assistant

Converts meeting transcripts and audio into structured summaries and action items with 90%+ accuracy using schema-constrained LLM outputs. Features real-time transcription, JWT and Google OAuth, PostgreSQL storage, and automated deadline reminder agents.

LLMReal-time TranscriptionPostgreSQL

StyleSavy – AI Fashion Consultant

Built an AI stylist analyzing 300+ outfits, boosting user engagement by 40%. Integrated YOLOS-Fashionpedia and FLAN-T5 for real-time clothing detection and recommendations via Gradio.

Gradio UIHugging Face TransformersYOLOS Object Detection

Predictive Analytics for County-Level Poverty in Florida

Visualized poverty and SNAP trends in 67 Florida counties, identifying a 15% poverty rise over 10 years. Built ARIMA and linear models, improving prediction accuracy by 22%.

RARIMAStatsmodels

Model Stealing Attacks on Multi-modal Model

Extracted surrogate models with 85%+ accuracy using CLIP pipelines on black-box AI systems. Found vulnerabilities in 70%+ of ResNet-18 test cases on CIFAR-10.

PythonNLPPyTorch

Warfarin Dosing Prediction Using Machine Learning

Built and tested multiple ML models (KNN, RF, SVR, FNN), improving dose prediction accuracy by 18%. Enhanced data quality and reduced errors by 25% through extensive cleaning.

PythonScikit-learnPandas

Stock Price Dashboard

Real-time dashboard in React + TypeScript with Tailwind displaying live quotes in a responsive, sortable table with sparkline charts. Integrated Finnhub and Alpha Vantage APIs with polling and caching.

React.jsTypeScriptTailwind CSS

Open Source Contributions

Projects I've contributed to in the open source community

TheHealthApp

A comprehensive healthcare application designed to improve patient care and streamline medical processes. Features include patient management, appointment scheduling, and health analytics.

React.jsNode.jsHealthcarePostgreSQLAPI Integration
View on GitHub

Education

My academic background and qualifications

Master's of Science in Artificial Intelligence

Saint Louis University (Jan 2024 - Dec 2025)

Saint Louis, MO, USA

Graduated with a Master's of Science in Artificial Intelligence with a focus on machine learning, deep learning, natural language processing, and AI applications in healthcare. Gaining expertise in cutting-edge AI technologies and research methodologies that drive innovation in intelligent systems and predictive analytics.

Bachelor of Engineering in Electronics and Telecommunications

A.C Patil College of Engineering (2017 - 2021)

India

Completed undergraduate studies in Electronics and Telecommunications with a strong foundation in engineering principles, mathematics, and analytical thinking that prepared me for advanced studies in artificial intelligence and machine learning.

Get In Touch

Have a project in mind or want to discuss opportunities? I'd love to hear from you!