Mercedes-Benz R&D · Bangalore · Open to opportunities
Software Engineer at the intersection of systems and AI
4+ years in production at Mercedes-Benz R&D — shipping C++ middleware, on-device AI systems, and CI/CD pipelines across 8 teams. Outside work, I build things: AI emergency responders, fan platforms, fully automated content pipelines. Curiosity is my default state.
Adaptive AUTOSAR
RAG · On-device LLMs
CI/CD Automation
C++17 · Python · Rust
Bronze Award
Impact at a glance
1,500+
Validation hours saved annually via CI pipelines
€5k+
Annual licensing costs eliminated
8
Teams using the CI infrastructure daily
5
Teams adopted the C++ SIL framework
300+
AUTOSAR test cases engineered end to end
Featured projects
Internal POC
AutoLog Expert Advisor
RAG · Gemma SLM · Sentence Transformers · Streamlit
On-device ECU log analysis using a locally-hosted small language model. Classifies log patterns across OTA, Diagnostics, and Powertrain domains — reducing manual analysis from hours to minutes.
Hackathon · GCP
Rapid Response AI
Gemini 2.5 Flash · TypeScript · Cloud Run
Analyses accident scenes, triages severity, and delivers structured first-response instructions in real time. Built during Google Cloud Run Hackathon using Gemini multimodal inference.
View on GitHub
Hackathon · MLB
MLB Fan Connect
Kotlin · GCP · TF-IDF · Recommendations
Fan engagement app built for the MLB Hackathon on Google Cloud. TF-IDF driven personalisation engine delivering real-time content recommendations based on user behaviour.
View on GitHub
Side project
ContentFlow
Google Cloud · Gemini · TTS · n8n · Python
Fully automated video publishing pipeline — from idea to live YouTube video with zero manual steps. Handles script generation, text-to-speech narration, multilingual subtitle generation, and auto-publishing.
Mercedes-Benz R&D
DriveContext
Real-time Telemetry · Geo-fencing · In-vehicle AI
Personalised in-vehicle recommendation engine using live vehicle telemetry and geo-fencing to surface driver-specific, location-aware suggestions. Nominated for a patent for algorithmic innovation.
Experience
Adaptive AUTOSAR Engineer
- Built a C++ SIL framework using Linux socket APIs to simulate ECU behaviour without hardware — adopted by 5 teams across Diagnostics, Security, and Communication
- Deployed GitLab CI sub-pipelines for full-release automation across 5 teams, saving 1,500+ annual validation hours
- Set up on-premise CI infrastructure eliminating CANoe license conflicts across 8 teams, saving €5,000/year
- Engineered 300+ test cases for the AUTOSAR communication stack covering SOME/IP, S2S, and IPC edge cases
- Built AutoLog Expert Advisor — an on-device RAG system using locally-hosted Gemma to classify ECU logs and auto-suggest fixes
- Scrum Master for the team — ran sprint planning, code reviews, and mentored 4 junior engineers
Graduate Engineer Trainee
- Built DriveContext — an in-vehicle recommendation engine using real-time telemetry and geo-fencing, nominated for a patent
- Developed a C# test framework via socket communication deployed across SOME/IP, Diagnostics, and Security teams
- Automated Python E2E pipelines with Jira Xray traceability — saving 800+ hours/year across 8 teams
Technical stack
Embedded & Middleware
Languages & Build
AI & Agentic Systems
DevOps & Cloud
Certifications & recognition
Bronze Award — CI/CD Infrastructure · Mercedes-Benz R&D