I am Phani M, a BTech Electronics and Communication Engineering student focused on building practical AI-powered software products. My work combines AI/ML engineering, backend development, full stack applications, and product-focused problem solving.
I enjoy building systems that move beyond basic demos: APIs that are structured cleanly, AI workflows that retrieve reliable context, frontend interfaces that users can actually interact with, and projects that can be explained confidently in interviews.
My current focus is on AI/ML applications, Generative AI, RAG systems, FastAPI backends, React frontends, and real-world product engineering.
| Currently Open To | Areas of Interest |
|---|---|
| Remote Internships | AI/ML Engineering |
| GenAI Projects | LLM Applications |
| Backend AI Roles | FastAPI + AI Workflows |
| Full Stack Internships | React + API Products |
| Freelance Projects | AI Tools and Automation |
| Domain | Proficiency | Details |
|---|---|---|
| Machine Learning | Intermediate | Classification, regression, evaluation, feature engineering, LightGBM |
| Generative AI | Intermediate | Gemini API, prompt engineering, structured AI workflows |
| RAG Systems | Intermediate | Document ingestion, chunking, embeddings, retrieval, citation-based answers |
| NLP | Intermediate | Text preprocessing, tokenization, transformers, sentiment analysis |
| Backend AI | Intermediate | FastAPI services, modular architecture, AI API integration |
| Full Stack AI Apps | Intermediate | React frontend connected with AI-powered backend APIs |
| Data Analysis | Intermediate | Pandas, CSV workflows, exploratory analysis, model output evaluation |
LexIntel — Legal Document RAG Assistant
An AI-powered legal document assistant that allows users to upload contracts or PDFs and ask questions with retrieved context and page-level source references.
| Category | Details |
|---|---|
| Stack | FastAPI, React, Gemini API, Qdrant, Python |
| Scale | Multi-page PDF ingestion with chunk-level retrieval |
| Performance | Retrieval-based answering with compact context generation |
| Security | API-based backend structure with controlled document flow |
| Impact | Helps users understand contracts faster with grounded answers |
| Repository | View Repository |
LexIntel was built as a practical RAG application with a modular backend. The system extracts text from PDFs, chunks the content, creates embeddings, stores them in a retrieval layer, and uses Gemini to answer questions using relevant document context. The project demonstrates backend AI architecture, retrieval pipelines, frontend integration, and real-world GenAI product thinking.
InternPilot-AI — AI Career Assistant
An AI-powered career guidance assistant designed to analyze user profiles and provide role-fit suggestions for internships and early career opportunities.
| Category | Details |
|---|---|
| Stack | FastAPI, LangChain, Python, LLM APIs |
| Scale | Modular backend with career analysis routes |
| Performance | Structured prompt flow for personalized output |
| Security | Environment-based API key handling |
| Impact | Helps students understand suitable career directions |
| Repository | View Repository |
InternPilot-AI focuses on using LLM workflows for career support. It includes backend routes, service layers, prompt handling, and career-fit analysis logic. The project helped strengthen my understanding of backend design, LangChain-style workflows, API debugging, and structured AI responses.
Product Review Sentiment Analyzer
A sentiment analysis application that classifies product reviews as positive or negative using a transformer-based NLP model.
| Category | Details |
|---|---|
| Stack | DistilBERT, FastAPI, React, Python |
| Scale | Review-level classification pipeline |
| Performance | High-confidence predictions on product review text |
| Security | API-based model inference flow |
| Impact | Useful for e-commerce review analysis and customer feedback understanding |
| Repository | View Repository |
This project connects a trained NLP model with a backend API and frontend interface. It demonstrates model inference, text classification, REST API design, and frontend integration for an end-to-end machine learning product.
Market Basket Analysis System
A recommendation-focused data mining project that discovers product association rules using transaction data.
| Category | Details |
|---|---|
| Stack | Python, Pandas, Apriori, Streamlit, FastAPI |
| Scale | Transaction-level product association analysis |
| Performance | Rule generation using support, confidence, and lift |
| Security | Local analytical workflow |
| Impact | Helps understand customer buying patterns |
| Repository | View Repository |
This project applies association rule mining to retail transaction data. It identifies product relationships and generates insights that can be used for recommendations, bundling strategies, and sales analysis.
Binance Futures Testnet Trading Bot
A Python CLI-based trading bot built for Binance Futures Testnet with support for market and limit orders.
| Category | Details |
|---|---|
| Stack | Python, Binance API, CLI, Logging |
| Scale | Command-line order placement workflow |
| Performance | Supports market and limit order execution |
| Security | Testnet-based development environment |
| Impact | Demonstrates API integration and trading automation basics |
| Repository | View Repository |
The trading bot helped me understand API-based automation, command-line argument handling, error logging, and external service integration. It was built for safe experimentation using Binance Futures Testnet.
Independent Projects
2025 — Present
Building AI-powered applications using FastAPI, React, machine learning models, and LLM APIs. Focused on developing projects that are practical, interview-ready, and aligned with real-world software engineering workflows.
Scope of Work
- Designed modular FastAPI backends for AI/ML applications
- Integrated LLM APIs into real product workflows
- Built React frontends for user interaction
- Created ML/NLP inference pipelines
- Worked with PDF processing, embeddings, retrieval, and prompt engineering
- Debugged backend issues, API errors, dependency conflicts, and deployment problems
| Recognition | Details |
|---|---|
| Hackathon Project Builder | Built and improved AI/ML projects for competitive and portfolio use |
| RAG Application Development | Developed LexIntel, a legal document assistant using retrieval-based AI |
| ML Competition Practice | Worked on ensemble-based prediction improvements and model evaluation |
| Full Stack AI Projects | Built multiple AI apps with backend APIs and frontend interfaces |
| GitHub Portfolio Development | Actively improving project documentation and profile presentation |
Learning:
- Advanced RAG pipelines
- LangChain and LLM application architecture
- Machine learning model evaluation
- Backend AI system design
Building:
- LexIntel legal document assistant
- AI career assistant workflows
- Full stack AI/ML portfolio projects
- Production-ready FastAPI services
Exploring:
- Prompt engineering
- Gemini API workflows
- Retrieval and citation-based AI systems
- AI agents and automation
Open To:
- Remote AI/ML internships
- GenAI internships
- Backend AI roles
- Full stack AI projects
- Freelance AI tools
