Ajinkya V. — AI Researcher

AI/LLM Specialist and System Software Engineer with a specialized background in the Autonomous Vehicle (AV) ecosystem. I bridge the gap between high-level AI model optimization and low-level system performance. My expertise is built on three core pillars: AI & LLM Optimization: Currently driving multilingual model performance for Meta’s MUPSL project at Innodata. I specialize in fine-tuning, instruction tuning, and high-fidelity dataset engineering for Llama-based models. Autonomous Systems: Previously at NVIDIA, where I validated perception data and HD mapping on the NVIDIA DRIVE platform. I have hands-on experience with ground truth generation (GTaaS), Deepmaps, and proprietary vision tools. High-Performance Computing: Proficient in C++ and CUDA, with a focus on GPU-accelerated vision pipelines, shared memory optimization, and real-time processing. I hold a PG-Diploma in Artificial Intelligence from CDAC and am B1-level certified in German. I am passionate about scaling the next generation of safe, autonomous, and intelligent software systems. Technical Stack: • Languages: C++, C, Python, Bash • Parallel Computing: CUDA, GPU Architecture, TensorRT • AI Frameworks: TensorFlow, Keras, PyTorch, LLM Fine-tuning • Tools: NVIDIA DRIVE, Docker, Linux Administration, OpenCV

Stackforce AI infers this person is a specialist in AI and autonomous systems with a strong focus on performance optimization.

Location: Pune, Maharashtra, India

Experience: 1 yr 7 mos