Headline

Backend Engineer | Java, Python | AI/RAG Systems | NITC

About

I’m a Software Engineer at CARIAD (Volkswagen Group) focused on backend engineering and AI-powered systems. I work primarily with Python, FastAPI, Java, Spring Boot, REST APIs, PostgreSQL, and scalable backend architectures. At CARIAD, I’ve worked on an AI-powered semantic video search platform covering 100+ hours of meeting recordings. My work spans backend services, automated speech-to-text and transcript indexing, semantic search, RAG pipelines, vector embeddings, and timestamp-accurate video retrieval. I also engineered CDN-integrated video delivery with adaptive bitrate streaming and chunked content delivery, helping reduce video playback latency by 40%. I enjoy building systems at the intersection of backend engineering and AI, particularly RAG, semantic search, vector retrieval, and LLM-powered applications. I also have experience with Node.js, Express.js, MongoDB, MySQL, Hugging Face Transformers, QLoRA, PEFT, Docker, Azure, Git, and CI/CD. I hold an M.Tech in Computer Science & Engineering from NIT Calicut (8.04/10) and a B.Tech in Computer Science & Engineering (7.95/10). I have also contributed to IEEE-accepted research on reliable higher-order MCQ generation using LLMs, focusing on prompting, fine-tuning, and bias-aware evaluation. Core areas: Backend Engineering • Python • Java • FastAPI • Spring Boot • REST APIs • RAG • GenAI • Semantic Search • Vector Embeddings • PostgreSQL • System Design • Microservices • DSA

Location

Bengaluru

AI stack

antigravity

codex

github copilot

hugging face

OpenRouter

Work experience

Trainee Software Engineer at CARIAD, Jul 2026 - Present · 1 mo, Bengaluru

Intern at CARIAD, Jan 2026 - Jun 2026 · 5 mo, Bengaluru, - Building an end-to-end AI-powered video search platform that enables teams to semantically search across meeting recordings. - Created a transcription pipeline to auto-generate transcripts from uploaded meeting videos, enabling full-text and semantic indexing. - Implemented vector embeddings on video transcripts and converted user queries into embeddings to retrieve the most contextually relevant video segments with precise timestamps. - Designed and developed the full-stack web application supporting video uploads, search, and results display with timestamp-level navigation. - Architected video delivery infrastructure using CDN-based streaming for scalable, low-latency video playback.

Education

Master of Technology - MTech, Computer science and Information technology, National institute of technology Calicut

Bachelor of Technology - BTech, Computer Science, Loknayak Jai Prakash Institute of Technology

Public links

leetcode: https://leetcode.com/anuragksng