Fastest Analytics & Search Database in the AI Era GitHub: github.com/apache/doris Discord: discord.gg/JemTbMKj
Sunnyvale, California
- Tweets 368
- Following 1,110
- Followers 388
- Likes 213
Join our webinar with @WeCloudData to build agent-facing analytics infrastructure. 👉 RSVP: velodb.io/events/agent-weclo…
Agents work differently from humans, so they bring a whole new set of requirements to the data infra that supports them.
1
2
60
Build a Snowflake alternative on AWS and get real-time analytics at a much lower cost
Results:
A fintech team moved TB-scale analytics from Snowflake to the AWS alt:
- 80% lower monthly cost
- 1-2 second ingestion latency
- up to 90x faster queries
🔗 velodb.io/blog/real-time-ana…
71
Come join us in San Francisco for the Full-Stack AI Hackathon and Mixer! Top prize get $270K in AWS credits 💸💸
The Executable World — full-stack AI hackathon & Demo & Networking
One day. The whole AI stack. $270K in AWS credits.
Sat 9/19 · 9 AM–8 PM · San Francisco
@VeloDB_IO @awscloud @TencentCloudEO @WorkBuddy_AI @AgentxAIgg @500GlobalVC @SSG_Accelerator @memories_ai PopPopTech
👇
93
Come meet the VeloDB and Apache Doris team at the AWS Well-Architected User Group meetups in Melbourne and Sydney later this month!
- Sep 23, Melbourne RSVP: meetup.com/en-au/aws-melbour…
- Sep 24, Sydney RSVP: meetup.com/aws-sydney-well-a…
64
We tested how Apache Doris, ClickHouse, Elasticsearch, and OpenSearch handle dynamic JSON fields in agent observability workloads.
Check out the result at 100 million rows:
🔗 More details in blog: velodb.io/blog/dynamic-json-…
1
106
We've added personalization steps to the VeloDB Cloud signup flow:
Kickstart your setup by selecting core workloads type: 1️⃣ Real-Time Analytics
2️⃣ Observability
3️⃣ RAG and GenAI
4️⃣ Lakehouse
🔗Try now, VeloDB Cloud offers 30-day trial and $300 credit: velodb.cloud/signup?utm_sour…
1
52
VeloDB Cloud now offers a native AI Assistant, right inside your dashboard.
We all have our favorite coding agents these days, but having a native AI assistant in a database dashboard can still be quite handy.
🔗 Try it out, 30-day trial and $300 credit: velodb.cloud/signup?utm_sour…
2
62
Meituan, a top food delivery app in Asia, rebuilt its multi-engine analytics stack with Apache Doris.
Results:
- Join queries improved by about 3x
- High-cardinality exact dedup improved 4x to 5x
- Query and ingest performance improved 20% to 40%
🔗velodb.io/blog/how-meituan-c…
121
Kwai, developer of Kling AI, rebuilt its A/B testing metrics production from Spark to Apache Doris.
This is the third Kwai use case. In the first two, Kwai replaced Clickhouse and Elasticsearch with Apache Doris in real-time serving for advertising.
🔗velodb.io/blog/from-spark-to…
107
Join VeloDB, dltHub, and Tower.dev for a live session on building the data stack agents actually need and love.
👉 RSVP: velodb.io/events/dlt-tower-2…
Demo:
Write an ETL job from scratch with dlt and Claude Code, deploy a dlt pipeline on Tower, query with VeloDB.
65
📣 📣 We have extended all VeloDB Cloud trials to 30 days with $300 free credits.
We understand picking a database is a key infrastructure decisions. We want you to make decisions based on real evidence and real workloads.
🔗 Try it out: velodb.cloud/signup?utm_sour…
70
VeloDB is partnering with Tower.dev to unify batch and real-time pipelines under one orchestration layer, so BI, dashboards, and AI agents all get served by a single engine.
We reworked the classic lambda architecture for AI and agents.
🔗velodb.io/blog/velodb-and-to…
71
Join VeloDB and GeoPITS for a live session on building a more unified real-time analytics stack.
👉 RSVP: velodb.io/events/geopits-202…
Live demo: streaming ingestion and CDC into VeloDB, plus federated queries across the lakehouse
📅 August 19, 4:00 PM IST | 6:30 PM SGT
43
We had a blast at the @dataengbytes in Sydney 🇦🇺
Two discussions we shared:
1️⃣ Production AI needs one engine, not a fragmented stack
2️⃣ Debug an AI agent needs real-time analytics, flexible JSON handling, and full-text search capabilities.
🔗Contact: velodb.io/apply/contact-us?u…
2
60
Thank you @techinasia for featuring VeloDB.
As our CEO Liam shared: “Models will keep changing, and the need for fresh, accurate context stays constant.”
#IMDA, #IMDAAccreditation
Even the fastest AI agents need fresh data. VeloDB helps enterprises unify analytics, search, and AI retrieval, giving agents better context to act on.
Learn how IMDA Accreditation supports firms like VeloDB: bit.ly/imda-velodb-tw
1
3
189
Join our webinar to see how Apache Doris 4.1 brings AI, search, and real-time analytics into one SQL engine. 👉 RSVP: velodb.io/events/doris-4-1-2…
Doris 4.1 also ranked first on ClickBench in cold query, and runs faster than 4.0 across standard benchmarks.
75
In our joint webinar, @KrantiParisa showed how @ApacheIggy delivered faster ingestion than Kafka:
- Throughput: 1,232 MB/s vs Kafka’s 371 MB/s
- Messages/sec: 4.81M vs Kafka’s 1.52M
- p99 latency: 2.5 ms vs Kafka’s ~40 ms
🔗Benchmark: github.com/velodb/velodb-dem…
5
319
StepFun built an agent observability platform in petabyte scale on Apache Doris @doris_apache .
@StepFun_ai is best known for its Step series LLMs (latest is Step 3.7).
Agent systems way are harder to debug than traditional microservices.
🔗velodb.io/blog/how-stepfun-b…
1
2
187
We tested Apache Doris, ClickHouse, Elasticsearch, and PostgreSQL on wide JSON workload
Setup:
10K unique JSON paths
No fixed schema
Each row fills 100 random keys (1% sparsity)
100M rows, 160 GB raw data
1K input files
16 cores, 64 GB RAM, SSD
🔗velodb.io/blog/beyond-10k-fi…
101
AI agents actually need real-time analytics more so than they need vector search.
Because many agents, like human analysts, deal with specific business, operational data first.
We broke down a query by the data layer capability required:
🔗 velodb.io/blog/why-ai-agents…
1
56