Vedant Kale

Final-year B.Tech CS student · Software Developer Intern

Pune, Maharashtra, India

@vedant_kale

LinkedIn GitHub

0 followers

🎓 B.Tech. in Computer Science & Engineering at Pimpri Chinchwad University (PCU) · Graduating 2027

About

Final-year Computer Science student and software developer intern targeting SDE roles in backend engineering, distributed systems, and AI systems. Built RAG and GraphRAG pipelines with hybrid search and retrieval evaluation, and has solved 150+ LeetCode problems.

Experience

Software Developer Intern

Karst Technologies Private Limited · Pune, India

Feb 2026 – Apr 2026

Contributed to development, debugging, and maintenance of application components. Performed software testing and validation to identify defects and support application reliability. Collaborated with the development team on API integration, database operations, and technical troubleshooting.

Education

2027

B.Tech. in Computer Science & Engineering

Pimpri Chinchwad University (PCU)

Computer Science · Aug 2023 – May 2027

Computer Science · Aug 2023 – May 2027

Skills

<3 months

C SQL RAG CSS3 Java Bm25 HTML5 Python FastAPI Node.js Next.js Three.js Pgvector Reacts Js Express.js JavaScript TypeScript SQLAlchemy Tailwind CSS React Native Microservices Hybrid Search Knowledge Graphs Crossencoder Reranking API Integration Understanding

Tools / apps / platforms

<3 months

Git Vite Linux Neo4j Redis Docker Ollama Pytest MongoDB Pinecone Jenkins CI Kubernetes PostgreSQL Amazon Web Services AWS

Projects

Adaptive Two-Tier Malware Detection

Python, Machine Learning, Random Forest, Evidential Deep Learning, API Security

Built a two-tier detection framework for SQLi, XSS, malware uploads, and RCE on 9,000 API requests. Added confidence-based escalation and feedback retraining, and evaluated on 10,000 CIC-IDS/EMBER records with high accuracy and low inference latency.
ArthaSEC – Financial RAG & Evidence Verification Engine

Python, FastAPI, BM25, Hybrid Search, SQLAlchemy

Built a financial retrieval engine over 32,000 SEC filing chunks, lifting Recall@10 from 42.6% to 73.3%. Added citation verification and deterministic calculations to validate metrics and flag unsupported figures, and optimized retrieval using BGE cross-encoder reranking.
Neuro-Adaptive GraphRAG Platform

Python, FastAPI, Neo4j, Pinecone, Cohere, RAGAS

Built a hybrid GraphRAG pipeline with Neo4j, Pinecone, and Cohere reranking, validated by 534 passing tests across 36 modules. Implemented RAGAS evaluation gates for retrieval quality and faithfulness with p95 latency monitoring, and developed a self-optimization framework balancing quality, latency, and feedback with a 15% canary rollout.

Courses & certifications

The Complete Full-Stack Web Development Bootcamp · Udemy · 2026
Data Structures and Algorithms Specialization · Amazon (via Coursera) · 2025
Software Design and Architecture Specialization · Coursera

🏆 Achievements & awards

Competitive Programming

LeetCode rating 1,480 with 150+ problems solved; CodeChef rating 1,824; CodeChef DSA rating 1,659.

Explore the Broxer community

💬 Q&A Ask about roles, salaries and interviews — and help others. 👥 Circles Join discussions on skills, companies and industries. 📰 Feed Follow updates and insights from the community.