Hello, I'm
SDE Intern @ Aptos · CSE-AIML @ PESU
Knight @ LeetCode · 3x Published
About Me
Hey folks! I'm Musadiq, a CSE-AIML graduate from PES University, Bangalore, currently interning as a Software Developer at Aptos India. I'm curious and driven, with interests spanning full-stack development, deep learning, and explainable AI.
I enjoy diving into new tech, picking up hands-on experience, and learning through real world applications from building production grade software to researching RAG systems and LLMs. Along the way, I love connecting with like minded individuals and sharing knowledge as we grow together in this ever evolving field.
I'm always on the lookout for exciting opportunities to upskill and make meaningful contributions.
Right now
SDE Intern @ Aptos
Bengaluru · Working on AI tools, Python, React
8.26
CGPA
8th
State rank, PUC
3
Papers published
Knight
LeetCode
Education
- PES University – B.Tech in Artificial Intelligence and Machine Learning (2022–2026)
- Grade: 8.26
- BGS PU College – Pre-University (PCMC, 2020–2022)
- Grade: 98.5% (State 8th Rank, 591/600)
- Vidyanidhi Public School – ICSE (2010–2020)
- Grade: 93%
Experience
Software Developer Intern
Aptos India Pvt Ltd
Jul 2026 – Dec 2026
Bengaluru, Karnataka · On-site
Software developer intern at Aptos, working on AI-powered tools for retail technology. Building with Python and React, and exploring AI agents as part of an engineering team.
Research Intern
CDSAML, PES University
Jan 2026 – May 2026
Bengaluru, Karnataka · On-site
Building an end-to-end Legal Judgment Prediction system for Indian Supreme Court cases using RAG and fine-tuned LLaMA 3.1 8B. Structured a dataset from 700+ cases and achieved BERTScore F1 of 0.85 on held-out evaluations.
Skills & Technologies
Programming Languages
Python
C++
C
Java
JavaScript
HTML
CSSML Libraries & Frameworks
TensorFlow
PyTorch
Pandas
NumPy
Scikit-learnNLTK
OpenCVHugging FaceWeb Development
Node.js
Express
React
Next.jsHTMLCSSDatabases & Tools
MongoDB
MySQL
Git
Docker
PyCharm
VS Code
GitHubGoogle Colab
KaggleMy Projects
Built a Knowledge Graph-enhanced RAG system for in-depth analysis of crime film narratives.
Read case study →Tech Used: Knowledge Graph, RAG, LLAMA 3.1, Neo4j , OllamaCreated multimodal meme virality predictor using Hypergraph Neural Networks for classification.
Read case study →Tech Used: HGNNs, CLIP, BERTBuilt Spotify-clone music streaming DBMS with user/admin roles, search & authentication.
Read case study →Tech Used: MySQL, React.js , JavaScriptBuilt Calendar microservice for teams to manage and share events efficiently.
Read case study →Tech Used: Docker, MicroserviceBuilt a client-server file transfer from scratch with chunking, SHA-256 checks and ACK/NACK retransmission.
Read case study →Tech Used: Python, TCP Sockets, FlaskBuilt BERT-based pipeline to classify, explain, and rewrite Turkish hate speech tweets.
Read case study →Tech Used: Python, Pandas, Transformers, SHAP
Stats & Activity
Research Papers
See all papers on Google Scholar →ORCID 0009-0007-6143-1102
Comparative Analysis of Traffic Accident Detection with Emphasis on Explainability of DL Models
Computer Vision
Utilized traditional ML and deep learning models including CNN, LSTM, and YOLOv8 to classify traffic accident images, with SHAP-based interpretability for enhanced decision transparency.
- Authors
- Mohammed Saqlain, Nishaan Padanthaya, K Musadiq Pasha, J O Shivnesh, Dhriti Rajesh Krishnan, K S Srinivas
- Venue
- International Journal of Advances in Electronics and Computer Science (IJAECS), Vol. 11, Issue 7, pp. 110-115
- Year
- 2024
- Dataset
- Accident vs. non-accident image dataset
- Methods
- CNN, CNN-LSTM, ANN, Random Forest, Decision Tree, SVM, YOLOv8
Beyond Loyalty and Betrayal: A Knowledge Graph-Enhanced RAG System for Crime Film Insights and Analysis
Natural Language Processing
Developed a Knowledge Graph and RAG system for analyzing crime films, enabling nuanced question answering and thematic exploration beyond factual retrieval.
- Authors
- K Musadiq Pasha, Tejas V. Bhat, K Bhavish Raju, Ayush Muralidharan, Bhaskarjyoti Das, D. L. Rameshwar
- Venue
- ICTIS 2025 proceedings, Springer LNNS book chapter (DOI 10.1007/978-981-96-8898-2_39)
- Year
- 2026 (published online 2 January 2026)
- Dataset
- The Godfather Part I script, 50 annotated QA pairs across 5 question types (public on Zenodo)
- Methods
- Neo4j knowledge graph, vector retrieval, LLaMA 3.1 via Ollama
HyperGNNs for Multi-Modal Classification and Severity Analysis of Neurodegenerative Disorders
Graph Neural Networks
Designed a Hypergraph Neural Network (HyperGNN) framework for multi-modal classification and severity prediction of neurodegenerative disorders, integrating heterogeneous clinical data to improve diagnostic accuracy and interpretability.
- Authors
- K Bhavish Raju, K Musadiq Pasha, Mohammed Saqlain, Nishaan Padanthaya, Jayashree R
- Venue
- 11th ICTIS, Bangkok, Thailand
- Year
- 2026
- Dataset
- MRI imaging with clinical tabular data
- Methods
- Hypergraph neural network with SHAP and Grad-CAM explainability
Certifications & Achievements
98.5%
PU College, State 8th Rank (591/600)
Knight
LeetCode, 1900+ contest rating, top 3.7%
1250+
LeetCode problems solved
Let's Connect
Always happy to connect, collaborate or chat about ML and SDE. Feel free to reach out!
