Linghua Jin
Cofounder & CEO of CocoIndex
Posts by Linghua Jin
-
Index Your Codebase for AI Agents with CocoIndex V1
Index a codebase for RAG and AI coding agents with CocoIndex V1 and Tree-sitter: language-aware chunking, embedding, and a live vector index in async Python.
-
CocoIndex V1 is Live!
CocoIndex V1 is live: a ground-up redesign of incremental data pipelines, built for AI engineers and agent builders shipping RAG, memory, and knowledge graphs.
-
CocoIndex joins the GitHub Secure Open Source Fund
CocoIndex joined the GitHub Secure Open Source Fund, hardening the AI data infrastructure developers depend on with threat modeling, CodeQL, and audits.
-
SEC EDGAR financial analytics with Apache Doris
A multi-format CocoIndex v1 pipeline that ingests SEC filings (TXT, JSON), scrubs PII, extracts topics, and powers hybrid search with Apache Doris.
-
Build a Self-Updating Wiki for Your Codebases with an LLM
Auto-generate documentation for every project in your codebase: a CocoIndex pipeline writes a wiki page per repo with an LLM, kept fresh as code changes.
-
Slides-to-speech: Bring your slide decks to life with narrated audio
Build a CocoIndex v1 pipeline that turns PDF slides into vision-generated speaker notes, local Pocket TTS narration, and searchable LanceDB records.
-
Extract patient intake forms with DSPy and CocoIndex
Extract Pydantic-typed structured data from patient intake forms using DSPy and CocoIndex: OCR vision models with incremental processing.
-
A knowledge graph from meeting notes that auto-updates
Build a self-updating Neo4j knowledge graph from meeting notes with CocoIndex v1: LLM-extracted decisions, tasks, and owners, with embedding-based dedupe.
-
What HackerNews is talking about: trending topics with an LLM
Rank trending HackerNews topics in Postgres with an incremental CocoIndex v1 pipeline: scrape threads and comments, extract topics with an LLM.
-
Extract HackerNews into Postgres with a custom source
Build a custom CocoIndex source for the HackerNews API: fetch threads and comments in async Python, extract topics with an LLM, keep Postgres in sync.
-
Extracting Intake Forms with BAML and CocoIndex
How to use BAML and CocoIndex to extract structured data from patient intake forms in PDF/Word with LLMs continuously for production.
-
AI-Native Data Pipeline - Why We Made It
Why the next wave of AI needs open-source, scalable, AI-native data infrastructure, and how CocoIndex is building the foundation for intelligent data pipelines.
-
Index PDF elements with mixed embedding models
Extract, embed, and store multimodal PDF elements (text with SentenceTransformers, images with CLIP) for unified semantic search with traceable metadata.
-
Bring your own data: Index any data with Custom Sources
Read data from any system with CocoIndex: in v1 a custom source is plain async Python plus processing components. No connector API to implement.
-
Trace search results back to source data
Define query handlers in CocoIndex and trace search results back to source data in CocoInsight to close the loop on indexing strategy.
-
Turn a Postgres table into a semantic index
Use an existing PostgreSQL table as a CocoIndex source: derive fields, embed each row, and store the vectors in Postgres with pgvector, incrementally.
-
Index PDFs, images, and slides with ColPali, no OCR
Build a unified visual document index from multiple file formats (including PDFs, images, and slides) using CocoIndex and ColPali. No OCR needed.
-
CocoIndex Changelog 2025-08-18
CocoIndex updates: production readiness, scalability, and reliability, plus more customization, native integrations, and multi-modal pipeline features.
-
Index Images with ColPali: multi-vector visual search
Index images with ColPali multi-vector patch embeddings and a Qdrant MaxSim collection using CocoIndex v1: incremental, live, in plain async Python.
-
Multi-Dimensional Vector Support in CocoIndex
CocoIndex natively handles typed multi-dimensional vectors, from simple arrays to multi-vector embeddings, unlocking multimodal AI pipelines at scale.
-
Custom Targets: export your data anywhere
Export CocoIndex data anywhere: implement a TargetHandler with a tracking record and an action sink; the engine handles diffing, syncing, and cleanup.
-
Index faces for visual search: your own Google Photos
Build a scalable face detection and recognition pipeline with CocoIndex: embed faces, structure for search, and export to a vector DB.
-
Index academic papers and extract metadata for AI agents
Index academic papers with CocoIndex: read the first page of each PDF, extract title, authors, and abstract with an LLM, and embed them for semantic search.
-
Introducing CocoInsight
Introducing CocoInsight, a data lineage and observability tool that lets you inspect, trace, and debug every step of a CocoIndex pipeline in real time.
-
Flow-based schema inference for Qdrant
CocoIndex sets up Qdrant collections automatically by inferring the target schema from your indexing flow: no manual config, vector sizes kept in sync.
-
CocoIndex + Kuzu: Real-time knowledge graph with Kuzu
Build a real-time knowledge graph with Kuzu as a native CocoIndex target: incremental updates, high-performance graph queries.
-
Incremental ETL on Amazon S3 with CocoIndex
Index Markdown from Amazon S3 into Postgres pgvector with CocoIndex v1: incremental processing re-embeds only the files and chunks that changed.
-
Image search in natural language with CLIP
Search a folder of photos by meaning with CocoIndex v1: CLIP embeds images and text into one vector space, the index runs live inside a FastAPI app, and vectors live in Qdrant.
-
How to build an index with text embeddings
Text embeddings 101: what they are, why you chunk and embed, and how to build a semantic search index with CocoIndex v1 and Postgres pgvector.
-
Story of CocoIndex, at 1k stars 🎉
The story of CocoIndex at 1,000 GitHub stars: the open-source engine that combines custom transformation logic with incremental processing for data indexing.
-
Build a product recommendation engine with LLM + Neo4j
An LLM extracts each product's taxonomy and complements; CocoIndex turns the labels into a Neo4j knowledge graph that answers 'bought this, also need…'.
-
Build Real-Time Knowledge Graph For Documents with LLM
Turn a folder of Markdown docs into a Neo4j knowledge graph: an LLM extracts subject-predicate-object triples, and CocoIndex keeps the graph in sync as the docs change.
-
Incremental Processing with CocoIndex
What incremental processing is, who needs it, and how CocoIndex keeps an index in sync with source changes through caching, lineage tracking, and change data capture.
-
Welcome to CocoIndex
Welcome to the official CocoIndex blog! We're excited to share our journey in building high-performance indexing infrastructure for AI applications.