VeloDB (Powered by Apache Doris)

Join our webinar to see how to retrieve better context for AI: velodb.io/events/doris4-… Agenda: 1️⃣ How Apache Doris 4.0's hybrid search lets your existing analytics database handle production AI workloads 2️⃣ @cocoindex_io + Apache Doris: AI agent pipelines, context engineering

Apache Doris 4.0 offers hybrid search and gives two levers to control that search layer: 1- What vector index you build 2- How you query it 🔗 Full deep-dive on hybrid search in Doris: velodb.io/blog/apache-do…

Our PM Wesley built an AI knowledge store with VeloDB, because he actually needed one. Part of his job: gather user feedback across GitHub, Jira, forums, and tickets to inform product decisions. The problem: data is siloed. Hybrid search saves it. 🔗velodb.io/blog/how-to-bu…

Postgres is a great place to start building (open source, reliable, pgvector...) But it wasn't built for real-time analytics. We have seen many teams hit this Postgres for analytics wall, and helped them fix it. The fix: PostgreSQL + Apache Doris. 🔗velodb.io/blog/postgresq…

🚀 CocoIndex 𝟎.𝟑.𝟑𝟒 𝐢𝐬 𝐡𝐞𝐫𝐞 — 8 releases since 0.3.27 featuring new target connectors, filesystem-level change detection, Python 3.14 free-threading, and smarter pipeline lifecycle management. We're super excited to build the amazing infrastructure projects together on the path of agents going to production in 2026! @doris_apache @VeloDB_IO @trychroma @falkordb @ladybugmem @lancedb 🔥Details in comment

VeloDB Cloud now supports one-stop data migration. The whole process runs through a web UI: connect your source cluster, select the databases and tables you want to move to VeloDB, configure concurrency, and go. No need to add scripts. 🔗velodb.io/blog/one-stop-…

ZTO Express delivers 35 billion+ packages a year across Asia. Their Trino + Hadoop analytics stack couldn't keep up. So ZTO used VeloDB on top of HDFS as a unified query engine, and used VeloDB internal table for real-time queries. 🔗velodb.io/blog/from-trin…

Come meet us at SCALE 23x in Pasadena, CA this week! The VeloDB team will be there: Booth #115 This year's sessions reflects something we think about a lot: how infra supports AI, cloud native, and real-time systems and supports open-source. 🎟️ register.socallinuxexpo.org/reg23/

SF Express, a global delivery company, runs a BI platform on top of an interactive lakehouse. They decided to replace #Presto with Apache Doris due to expanding workload. Result: - 100% Presto workloads migrated - 3x faster queries - 48% cost savings 🔗velodb.io/blog/sf-techno…

Thank you, Bengaluru! 🇮🇳 We wrapped up an incredible Saturday at the Apache Kafka meetup with OLake and Confluent. Key Takeaway: Real-time analytics is becoming the infra foundation for RAG and AI apps. Thank you to everyone who showed up and stayed for the great discussions!

This Saturday in Bangalore, come meet us in the Apache Kafka meetup! See how Apache Doris integrates with Apache Kafka, with native Kafka Connector that handles continuous ingestion from Kafka topics into Doris, supporting multiple formats (JSON, Avro) 🔗velodb.io/events/bangalo…

Last call 📢 Our webinar on real-time context engineering is this Thursday. If you're building RAG pipelines, AI code agents, or any AI tool that needs fresh and real-time context, this is for you. 👉RSVP: velodb.io/events/cocoind…

A Web3 data team replaced their Apache Doris + Trino + Spark stack with just Apache Doris and Apache Paimon. Results: - Lake queries 2x faster than #Trino - ETL 5x faster than Spark - Data freshness shorted from T+1 to minutes - 40% lower storage costs 🔗velodb.io/blog/apache-do…

Vector search is great but applying it on large dataset can be expensive. See how #ByteDance utilizing Apache Doris 4.0 hybrid search capability in a search project with 1B+ #vectors: They're combining vector search with full-text search and structured filtering.

Apache Doris 4.0 is bringing vector search and AI functions in one engine. VeloDB, @lyzr__ai, and AI Circle are hosting a meetup on February 21 in Hyderabad. Come see what Apache Doris looks like in practice. 🔗 RSVP: velodb.io/events/hyderab…

In Web3, you needs two things at once: 1- A stream processor that can keep up with the massive real-time data of crypto → That's Flink 2- An OLAP engine that can handle high-concurrency, real-time and historical queries → That's Apache Doris 🔗velodb.io/blog/build-rea…

Advance.AI, Singapore AI and fintech company, cut 50% of their log ananalytics costs by switching from OpenSearch to Apache Doris. They chose Apache Doris after evaluating it with ClickHouse, Loki, and others.

CocoIndex and VeloDB form an AI context stack that brings real-time context updates and retrieval. They give your RAGs and agents fresh context and good memory. 1. @cocoindex_io is the transformation layer 2. VeloDB is the retrieval layer 🔗Blog: velodb.io/blog/the-real-…

A top crypto exchange built their entire analytical platform on VeloDB to handle 80 million users and $10 billion in daily trading volume. 3 layers: 1. Stream processing > Flink + Redis 2. Real-time analytics > VeloDB 3. Lakehouse > VeloDB + Paimon 🔗velodb.io/blog/how-a-top…

Join our webinar with @cocoindex_io to learn how to keep context fresh for #RAG, code agents, and AI tools 👉 RSVP: velodb.io/events/cocoind… Live demo: Use CocoIndex + VeloDB in a codebase, get continuous index updates as code changes >> real-time context stack for AI