Manav Gupta

Pluto Labs: Solving Problems that really matters (bootstrapped) > Created jail break LLMs for Survival and Medical Scenarios (S5 Buildspace) > Currently trying to solve the "Tokenisation" Problem of AI Agents

Google me, I am Tensorboy

- Created few datasets form scratch - Learned Data Engineering - Won few competitions

- Improved my data intuitions - Increased my reasoning capabilities for any solution - Helped me became consistent

- Leans the basics of ML - Worked on my problem solving and experimentation skills

- Playing around with Diffusion Models - So called "director" of AI models - SDXL and Flux

- Building the CDSS system, powered by data - Currently working on TensorRT, VGT Models & Triton server - Bench-marking pipelines and algorithms - Understanding the back-end architectures used to deploy LLMs on scale.

- Deploying open-source LLM pipelines - Improved my documentation reading skills (Figuring out) - Fine tuned my first ever bert/ diberta models for sentiment detection - Informed and advised startups ways to save cost using open source

- Architecting Pluto - LLM solution you actually need, in all this chaos

-Developed affordable RAG pipelines using open-source, cutting costs 90% -Boosted data retrieval efficiency 65%+ -Explored Sentence Transformers & Vector Embeddings -Built frameworks benchmarking RAG models

- Read Papers, a lot of them - Improved my NLP skills - Actually understand the enterprise level deployment - Finally, enhanced tech intuition for the solutions

I resigned, just a bunch of immature guys, fighting over credit. Internal Politics is harmful for the health. - Organised sessions on ML - Promoted Data Science as an career option - Solved doubts and guided juniors in the same feild