Integrated M.Tech in Computer Science Engineering | University of Hyderabad

A motivated software engineer with a strong academic foundation and hands-on experience in full-stack web development, system architecture, and artificial intelligence. I enjoy combining analytical puzzle-solving with UI/UX principles to build efficient, user-centric applications.

I am currently dedicating my time to a few long-term engineering initiatives:

• TUI-do: Developing a terminal-native Kanban board for the FOSS Hack event.
• Enterprise Banking System: Architecting a comprehensive, secure banking platform as a semester-long software engineering endeavor.
• Code-Smell AI: Training a machine learning model designed to detect "structural smells" in codebases.

• Languages: C, Java, Python, SQL, JavaScript, HTML/CSS, PHP
• AI & Machine Learning: LangChain, LangGraph, FAISS, Vector Embeddings, Groq, HuggingFace
• Web Technologies: ReactJS, Node.js, Puppeteer
• Databases & Tools: MongoDB, MySQL, Git, Docker, NetworkX
• Core Concepts: Data Structures & Algorithms, Object-Oriented Programming, Operating Systems, Computer Architecture, Compiler Design

Structural Centrality Scoring (SCS) — SmartCom 2026

• Introduced a metric-driven workflow to transform static call graphs into actionable architectural insights.
• Developed a novel metric combining Betweenness Centrality and PageRank to identify architectural hotspots.
• Built a Python/NetworkX toolchain to filter architectural noise by 30%, differentiating between runtime overhead and structural importance.

• F1 Insight Agent: Architected a specialized AI agent using LangGraph and Groq to perform multi-step reasoning over Formula 1 datasets. Integrated external APIs for real-time telemetry and race statistics.
• Multi-Source RAG Pipeline: Developed a Retrieval-Augmented Generation system using FAISS and Hugging Face embeddings to enable semantic search over unstructured documents, significantly reducing LLM hallucination rates through recursive character splitting.
• Electricity Bill Generator: Engineered a full-stack digitization platform for TSSPDCL billing cycles. Implemented a 3-tier RBAC system with automated logic for rolling arrears and late fines, plus a custom CSS engine for thermal-style receipts.
• Headless Web Scraper: Built a Puppeteer-based headless browser scraper to extract multi-category data, supported by custom REST API endpoints for triggering scrapes and retrieving sanitized CSV data.

• Email: kpradyun18@gmail.com
• Location: Hyderabad, Telangana, India