Yatharth Kapadia

Yatharth Kapadia⁠

About

I build low-latency retrieval for voice apps and high-throughput model serving systems.

Experiences

  1. Artificial Intelligence Engineer at CogniwareAIMake big AI models fast and cheap to run

    Most of it is pulling throughput from the GPUs you have and routing each request to the model that should handle it.

  2. MossShipped Moss features downloaded 300K times

    We build sub-10ms semantic search for conversational AI, shaving at least 200-300ms off voice AI stacks.

  3. Hackathon Judge at MossJudged 2 YC hackathons, 200 builders each

    What surprised me: how many teams reached for Moss as their retrieval layer, and how much could ship as-is.

  4. WritingWrote 2 guides to running LLMs in production

    One covers GPU limits, VRAM and quantization. The other, Python threading and caching that cut API costs up to 90%.

  5. Graduate Teaching Assistant (E-511: Machine Learning for Signal Processing) at Indiana University Luddy School of Informatics, Computing, and EngineeringResearched LLM hallucinations, taught AI

    At Kelley I probed Hugging Face LLMs for security holes. I also led lectures for a 70-student AI course.

  6. InPersonaBuilt a knowledge graph that answers as me

    It runs RAG pipelines on a fine-tuned LLaMA 3.2 11B, with an agent watching for prompt injections.

  7. Machine Learning Intern at EloCut a speech-to-speech pipeline below 800ms

    I streamed Whisper STT and TTS over WebSocket, and helped fine-tune Llama 3.1 to rely less on ChatGPT's API.

  8. Software Developer Intern at Quidich Innovation LabsPredicted cricket throws to 95% accuracy

    At Quidich I forecast ball release from player coordinates. At IDeaS, automation saved 500+ hours of manual work.

Follow Yatharth for more...

LinkedIn

© 2026 Delphi · Terms · Privacy

Yatharth Kapadia

Yatharth Kapadia⁠

About

I build low-latency retrieval for voice apps and high-throughput model serving systems.

Experiences

  1. Artificial Intelligence Engineer at CogniwareAIMake big AI models fast and cheap to run

    Most of it is pulling throughput from the GPUs you have and routing each request to the model that should handle it.

  2. MossShipped Moss features downloaded 300K times

    We build sub-10ms semantic search for conversational AI, shaving at least 200-300ms off voice AI stacks.

  3. Hackathon Judge at MossJudged 2 YC hackathons, 200 builders each

    What surprised me: how many teams reached for Moss as their retrieval layer, and how much could ship as-is.

  4. WritingWrote 2 guides to running LLMs in production

    One covers GPU limits, VRAM and quantization. The other, Python threading and caching that cut API costs up to 90%.

  5. Graduate Teaching Assistant (E-511: Machine Learning for Signal Processing) at Indiana University Luddy School of Informatics, Computing, and EngineeringResearched LLM hallucinations, taught AI

    At Kelley I probed Hugging Face LLMs for security holes. I also led lectures for a 70-student AI course.

  6. InPersonaBuilt a knowledge graph that answers as me

    It runs RAG pipelines on a fine-tuned LLaMA 3.2 11B, with an agent watching for prompt injections.

  7. Machine Learning Intern at EloCut a speech-to-speech pipeline below 800ms

    I streamed Whisper STT and TTS over WebSocket, and helped fine-tune Llama 3.1 to rely less on ChatGPT's API.

  8. Software Developer Intern at Quidich Innovation LabsPredicted cricket throws to 95% accuracy

    At Quidich I forecast ball release from player coordinates. At IDeaS, automation saved 500+ hours of manual work.

Follow Yatharth for more...

LinkedIn

© 2026 Delphi · Terms · Privacy

By using this service, you agree to the Terms of Service and Privacy Policy.