Headline

Senior AI Engineer | LLMs, AI Agents & Generative AI | Full-Stack ML | IIT Madras

About

Senior AI Engineer with 5+ years building end-to-end AI systems — from data curation, model training and fine-tuning to inference optimization and production deployment. Currently at Livepeer working on real-time video diffusion models and AI infrastructure. Previously at Sprinklr (4+ years) building production ML at scale — LLM agents, NLP, computer vision, and ML infrastructure serving enterprise clients. I go beyond black-box usage — I understand the full stack from model architecture to GPU serving and like to build systems that are fast, reliable, and cost-efficient. Areas: LLMs, AI Agents, Generative AI, Diffusion Models, NLP, Computer Vision, Model Training & Fine-Tuning, Inference Optimization, TensorRT, PyTorch, Transformers, MLOps, Distributed Systems Open to Senior AI/ML Engineer, Forward Deployed Engineer, Applied AI Scientist, or Founding Engineer roles.

Location

India

AI stack

codex

cursor

opencode

Work experience

Senior AI Engineer at Livepeer Studio, Nov 2024 - Present · 1 yr 10 mos, Remote, Generative AI, Video and Image Diffusion models, model training & fine-tuning, TensorRT, inference optimization, quantization, ControlNets, AI infrastructure

Lead Product Engineer | AI Team at Sprinklr, Apr 2023 - Nov 2024 · 1 yr 8 mos, Led ML infrastructure — made in-house LLM serving 10x cheaper than GPT-3.5/4 and built frameworks that boosted GPU throughput 4x across all services. LLM agents, inference optimization (TensorRT, vLLM, speculative decoding), GPU serving, ML infrastructure, throughput & latency optimization

Senior Product Engineer | AI Team at Sprinklr, Aug 2022 - Apr 2023 · 9 mos, Worked on production NLP services as part of small (2-3 person) teams — including a next-word suggestion engine serving under 100ms, a context-aware response generation system on GPT-2, grammar correction service built with T5 model. Smart Compose, Smart Reply, Voice Agent Assist, GPT-2, wav2vec2, NLP, MQTT, encoder-decoder models (T5), real-time serving (MQTT, gRPC)

Product Engineer | AI Team at Sprinklr, Aug 2020 - Aug 2022 · 2 yrs 1 mo, Contributed to a couple of production ML services across NLP and computer vision — including a logo detection service outperforming Google Cloud Vision across 5000+ logos. Computer Vision (YOLOv5)

Summer Intern at American Express, May 2019 - Jul 2019 · 3 mos, Bengaluru Area, India, Developed a robust and efficient API using Spring Boot, facilitating seamless synchronization of data between a bunch of dynamic web pages and an internal database to eliminate manual effort, ensuring data consistency and integrity.

Summer Intern at Godot Media, May 2018 - Jul 2018 · 3 mos, Bengaluru, Karnataka, India, Developed a search functionality built on top of Elastic Search, including auto-suggest features to enhance user navigation on a website. Additionally, designed a personalized recommendations system on a large database by implementing data clustering and ranking algorithms, boosting user engagement and delivering tailored experiences.

Education

Bachelor of Technology - BTech, Computer Science, Indian Institute of Technology, Madras

Public links

github: https://github.com/varshith15