Yashvardhan Thanvi's profile on Product Hunt

About

CS/AI undergrad at BIT Mesra, currently an AI Research Intern at DRDO SAG Lab working on NLP-driven pipelines, RAG systems, and LLM-based telemetry. Global Finalist at NASA Space Apps 2025 and a Google Student Ambassador. I build practical AI tools that solve real engineering pain points — most recently LLMSlim, an open-source Python package that cuts LLM token costs by 40-70% through offline semantic prompt compression, with a guaranteed 100% instruction retention. Always keen to connect with builders and researchers working on efficient, production-grade LLM systems. Connect with me on LinkedIn: linkedin.com/in/yashvardhanthanvi Explore LLMSlim: llmslim.app

Links

Twitter

Badges

Tastemaker

Gone streaking

Maker History

LLMSlimContract-safe context tooling for LLM tool schemas.

🎉

Joined Product HuntJuly 15th, 2026

Forums

•

6d ago

LLMSlim - Tool-aware context, contract-safe.

v0.4.0 adds contract-safe tooling for MCP, OpenAI function, Anthropic, and generic tool schemas: provider-aware adapters, deterministic canonical JSON, SHA-256 fingerprints, exact-equivalence checks, and schema inspection. Raw execution schemas remain authoritative. Tool retrieval is research-only and is never used for authorization or execution.

•

1mo ago

LLMSlim v0.3.0 - Cut LLM token costs 40-70% with offline prompt compression

v0.3.0 ships hybrid prompt compression: offline TF-IDF extraction pre-prunes context, then an optional LLM rewrite pass semantically optimizes what remains. Result: 40-70% fewer tokens billed, sub-30ms CPU latency, and 100% retention of system directives, code blocks, and JSON schemas. Works with OpenAI, Anthropic, Gemini, LangChain, LlamaIndex, Ollama and more. Zero dependencies. Pure Python. pip install llmslim

•

2mo ago

LLMSlim - Semantic prompt compression that never drops instructions.

Open-source Python package cutting LLM token costs by 40–70% in 1 line of code. Surgically compresses prompts, RAG document contexts, and multi-turn chat logs with guaranteed 100% instruction fidelity.