50+ merged PRs across YC-backed AI infrastructure projects
Unsloth AI
Contributor · 70k+ GitHub stars · unslothai/unsloth
Contributed 50+ merged pull requests to this open-source LLM fine-tuning and inference platform. Improved Unsloth Studio UX (touch sidebar support, model presets, token counts, Colab embeds, capability detection fixes), enhanced the cross-platform installer and GPU stack (Vulkan/ROCm routing, distro detection, dependency pinning), and fixed model-loading reliability (offline tokenizer loads, HF cache recovery, Deep Research context-window safety).
RAGFlow (InfiniFlow)
Maintainer — MinerU PDF Parser · 90k+ GitHub stars · infiniflow/ragflow
Invited into the InfiniFlow GitHub organization as dedicated maintainer of RAGFlow's MinerU PDF/layout parser — triage, review, and follow-through on issues and PRs. Implemented MonkeyOCR as a dedicated PDF backend on top of the MinerU parser class. Fixed dataset Artifacts Tree silently failing on navigation API errors, and stopped agent-canvas autosave from wiping working pipelines by persisting an empty graph.
Dify
Contributor · 150k+ GitHub stars · langgenius/dify
Contributions to this open-source LLM app development platform, working across agent tooling and workflow orchestration.
LiteLLM
Contributor · 55k+ GitHub stars · BerriAI/litellm
Contributions to this unified LLM API/proxy layer used to call 100+ LLM providers in the OpenAI format.
Mastra
Contributor · 27k+ GitHub stars · mastra-ai/mastra
Contributions to this TypeScript AI agent framework, working on agent tooling and framework internals.
AnythingLLM
Contributor · 64k+ GitHub stars · Mintplex-Labs/anything-llm
Contributions to this full-stack application for turning documents into a chat interface with any LLM, working on model management and API security.