Aditya Prakash — Research Engineer & Software Engineer

Background

WHO AM I

A Computer Science student driven by curiosity about how things work at a fundamental level.

I'm Aditya Prakash, pursuing Computer Science and Business Systems at CVR College of Engineering. My academic journey has been defined by a genuine fascination with computer science fundamentals, the kind that makes you want to understand not just how to use a tool, but why it works the way it does.

My research interests span operating systems, compiler design, natural language processing, and machine learning. I enjoy the intersection of theory and practice: taking mathematical ideas and turning them into reliable, efficient software.

Beyond coursework, I contribute to open-source projects like SymPy, where my merged PR #29172 improved core mathematical error handling. I believe the best way to learn is to build, and the best way to build is to understand the foundations.

When I'm not coding, I'm reading research papers, exploring algorithm design, or preparing for GATE CS, always looking for the next problem worth solving.

🎓

CVR College of Engineering

Computer Science and Business Systems (CSBS)

CGPA: 8.7 2024 – 2028 Hyderabad, India

Relevant Coursework

Operating Systems Compiler Design Algorithms Data Structures DBMS Machine Learning Computer Networks Discrete Mathematics Theory of Computation

Interests

Research Interests

Areas where I direct my curiosity and study, from theory to application.

⚙️

Operating Systems

Process scheduling, memory management, kernel design, and concurrency primitives.

🔣

Compiler Design

Lexical analysis, parsing, semantic analysis, IR generation, and optimization passes.

💬

Natural Language Processing

Text representation, transformers, tokenization, and computational linguistics.

🧠

Large Language Models

Pretraining, fine-tuning, alignment, inference optimization, and evaluation.

📈

Machine Learning

Supervised learning, model explainability, probabilistic methods, and neural architectures.

🔗

Graph Algorithms

Graph traversal, shortest paths, network flows, and graph neural networks.

🏗️

Software Engineering

System design, architecture patterns, API design, and maintainability at scale.

🔍

Explainable AI

SHAP, LIME, attribution methods, and building trustworthy ML systems.

Projects

Engineering Work

Systems built with intention, from research prototypes to production-grade software.

Research · Security

Kavach (PhishXray)

A real-time phishing detection system combining a Chrome Manifest V3 extension with a FastAPI backend powered by fine-tuned RoBERTa. Features a Risk Fusion Formula aggregating signals from NLP, URL heuristics, and visual cues for explainable threat verdicts.

Problem: Phishing attacks exploit cognitive bias under time pressure. How do you build a system that both detects threats accurately and explains its reasoning to the user?

FastAPI RoBERTa SHAP LIME Manifest V3 Python

View on GitHub

Software Engineering · AI

SkillBridge

A resume skill gap analyzer that parses uploaded resumes, identifies missing skills relative to a target job description, and generates an AI-powered personalized learning roadmap with recommended resources.

Problem: Job seekers often don't know specifically what skills they lack. SkillBridge gives them actionable, targeted guidance — not generic career advice.

Python NLP Scikit-Learn AI Recommendations

View on GitHub

Academic · Systems Programming

Hotel Management System

A console-based hotel management system written in C, implementing room booking, check-in/check-out workflows, billing calculation, and persistent data storage using file I/O. Demonstrates strong fundamentals in C, memory management, and systems-level thinking.

Problem: Design a complete data management system using only C and the standard library — no frameworks, no shortcuts. Pure systems programming from the ground up.

C File I/O Data Structures Console UI

View on GitHub

Open Source

Contributions

Going beyond courses: reading source code, understanding maintainer conventions, and shipping real fixes.

Σ

SymPy — PR #29172

Core Mathematical Error Logic · python/sympy

✓ Merged

🔍

Problem Identified

While using SymPy for a numerical analysis assignment, discovered an edge case where certain mathematical error conditions produced incorrect or misleading results in the core symbolic engine.

💬

Issue & Discussion

Filed a detailed issue with a reproducible minimal example. Engaged with maintainers to understand the expected behavior, SymPy's contribution guidelines, and the appropriate fix strategy.

$ git clone https://github.com/sympy/sympy.git
$ cd sympy && git checkout -b fix/error-logic-29172

⚙️

Implementation

Implemented the fix in the core mathematical error handling module. Added comprehensive tests to prevent regression, following SymPy's strict testing conventions and ensuring backward compatibility.

✅

Merged

PR reviewed by core maintainers and merged into SymPy's main branch. This contribution is now part of every SymPy release, used by researchers and engineers worldwide.

View PR #29172

Journey

Timeline

A record of milestones: academic, engineering, and research.

24

2024

Started CSBS at CVR College of Engineering

Began the Computer Science and Business Systems program, diving deep into algorithms, data structures, and the theoretical foundations of computing.

25

2025

AI XPO — Hackathon Runner-Up

Placed Runner-Up at AI XPO, presenting Kavach (PhishXray), a real-time phishing detection system combining NLP, explainability, and a Chrome extension.

25

2025

SymPy PR #29172 Merged

Became a contributor to SymPy, one of the most widely used computer algebra systems in the Python ecosystem. Core mathematical error logic fix shipped to users worldwide.

→

Present

Building · Learning · Contributing

Developing Kavach further, studying Operating Systems and Compilers in depth, exploring research paper writing, and continuing open-source contributions.

Toolkit

Skills

Tools, languages, and concepts I work with regularly.

Programming Languages

Python Java C SQL

Libraries & Frameworks

NumPy Pandas Scikit-Learn FastAPI Transformers

Developer Tools

Git GitHub Docker VS Code Linux

Databases & Soft Skills

MySQL Research Problem Solving Analytical Thinking Communication

Recognition

Achievements

🏆

AI XPO Runner-Up

Hackathon · 2025

Placed Runner-Up at AI XPO for Kavach (PhishXray), a real-time phishing detection system combining ML explainability with a Chrome extension.

🔧

SymPy Contributor

Open Source · PR #29172

Merged contribution to SymPy's core mathematical error logic. Code now ships in every SymPy release to researchers and developers worldwide.

📄

Research Publications

In Progress

Actively working toward research paper submissions in NLP and explainable AI. Target venues include undergraduate research tracks of major CS conferences.

Writing

Notes from the Lab

A collection of articles documenting exploration of Computer Science: NLP, Operating Systems, Algorithms, Compilers, AI, and Software Engineering.

NLPTransformers

Understanding Attention from First Principles

A bottom-up derivation of self-attention, from the intuition of similarity scoring to the full scaled dot-product formulation. With worked examples in NumPy.

Coming Soon· ~12 min read

CompilersTheory

How Compilers See Your Code: A Parse Tree Walkthrough

Walking through lexical analysis, context-free grammars, and recursive descent parsing, illustrated with a small expression evaluator built from scratch.

Coming Soon· ~10 min read

Open SourceSymPy

My First Open Source Contribution: What I Learned

A walkthrough of contributing to SymPy, reading a large unfamiliar codebase, communicating with maintainers, writing tests, and getting a PR merged.

Coming Soon· ~8 min read

MLExplainability

SHAP vs LIME: When Explanations Disagree

A practical comparison of SHAP and LIME for model explainability, when they agree, when they don't, and what that tells you about your model and data.

Coming Soon· ~9 min read

Now