Skip to content
View Prathmeshkangane's full-sized avatar

Block or report Prathmeshkangane

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
Prathmeshkangane/README.md

Hi, I'm Prathmesh πŸ‘‹

πŸŽ“ B.E. CSE (AI & ML) '26 | University of Mumbai (Saraswati College of Engineering)
πŸ€– AI/ML Engineer | LLMs & RAG | NLP | Deep Learning | Data Science
πŸ“ Mumbai, India | Open to Work (fresher, open to relocation across India)

I build end-to-end AI systems, from model training to deployed apps. My work covers speech, NLP, and retrieval-augmented generation (RAG).


πŸš€ Featured Projects

🎀 AI Mock Interview Platform

Production-grade interview coach with real-time speech recognition, emotion tracking, STAR-method scoring, filler-word detection, WPM analysis, and PDF reports.
Stack: React, TypeScript, FastAPI, Gemini 2.5 Flash, custom DistilBERT classifier (Hugging Face)
πŸ”— Model Β· Code

πŸ“š AI Research Assistant (RAG over ArXiv)

Semantic search and question answering over 10,000 ArXiv papers using vector retrieval and LLM-generated answers. Containerized with Docker Compose (app + MySQL).
Stack: Flask, FAISS, Sentence Transformers, Gemini, MySQL, Docker
πŸ”— Code

🎧 Speech Emotion Recognition

Stacked LSTM classifier on MFCC and spectrogram features across 7 emotion classes, trained on RAVDESS + TESS. Reached ~81% test accuracy (macro F1 0.81).
Stack: TensorFlow/Keras, librosa, Streamlit
πŸ”— Code


πŸ“„ Publication

Multimodal AI System for Automated Interview Analysis
International Journal of Emerging Developments in Engineering and Research (IJEDR), Vol. 14, Issue 2, April 2026


πŸ› οΈ Skills

Languages: Python SQL
ML / DL: TensorFlow PyTorch CNN RNN LSTM Transformers scikit-learn
GenAI: LLM APIs (Gemini) RAG FAISS Sentence Transformers Hugging Face
NLP / Vision / Audio: NLP Computer Vision Speech Processing
Data: MySQL NoSQL Kafka Data Analysis Data Visualisation
Backend & Tools: Flask FastAPI Docker Git AWS Streamlit


πŸ’Ό Internships & Experience

Role Organisation Period
AI-Based Image Transformation Tool for Cartoon Effect Generation (Internship 6.0) Infosys Springboard Feb – Apr 2026
Software Development Engineer (SDE) Intern, remote Bluestock Fintech Jan – Feb 2026
Emerging Technologies (AI & Cloud) internship with IBM SkillsBuild & IBM Cloud, in collaboration with AICTE Edunet Foundation Jul – Aug 2025
AI/ML internship via AICTE/Elevate Microsoft 2025

πŸ… Certifications

AI & Cloud

Programming & Software Engineering

  • Udemy: 100 Days of Code, The Complete Python Pro Bootcamp (56.5 hrs, Jul 2025)
  • Infosys Springboard: Object Oriented Programming using Python (Jun 2025)
  • Infosys Springboard: Software Engineering and Agile Software Development (Jun 2025)

Databases

  • Infosys Springboard: Database Management System, Part 1 (Jun 2025)
  • Infosys Springboard: Introduction to NoSQL Databases (Jun 2025)

Data Analytics (Forage job simulations)

  • Deloitte: Data Analytics Job Simulation (Jun 2025)
  • Tata: Data Visualisation, Empowering Business with Effective Insights (Jun 2025)

🧩 Problem Solving

  • Striver's A2Z DSA Sheet: 106/455 solved

πŸ“« Connect

LinkedIn Gmail Hugging Face

Pinned Loading

  1. interviewAi interviewAi Public

    TypeScript 1

  2. AI-Research-Assistant AI-Research-Assistant Public

    HTML

  3. Emosense Emosense Public

    Python