Real-Time Analytics & Search DB

The unified SQL database for engineers and Al agents.

Start Free Trial

Contact Sales

Combined search and analytics for a simplified data platform

80%

Storage

Throughput

Unified

Cost-effective json analytics without sacrificing performance

<2 Sec

Search Latency

10GB/s

Throughput

70%

Storage Cost

Unifying Analytics on Lakehouse

4 to 2

Databases

3-5x

Query Latency

Concurrency

Real-time AB testing over tens of billions of daily events

Throughput

Concurrency

Data Freshness

Gartner Data & Analytics Summit in Sydney

VeloDB is exhibiting at the Gartner Data & Analytics Summit 2026! Come talk data with us at Booth 113, and don't miss our session at Theater 1 on June 17 at 3 PM.

Learn more

10,000+ Enterprises Trust Apache Doris and VeloDB

REAL-TIME

The Fastest End-to-End Engine from Ingestion to Insight

1s Latency

Streaming Ingestion

Supports high-frequency, second-level streaming ingestion from both Kafka and Database CDC.

< 1min Refresh

Incremental Transform

Replaces batch processing with incremental materialized views that refresh complex transformations in minutes.

10x Speedup

Blazing-Fast Analytics

10x on complex analytics . Under concurrency, it maintains a consistent 10ms response time across PB-scale datasets.

MPP

Cost-Based Optimizer

Vectorized Execution

Inverted Index

Skip Index

Unique Model

Materialized View

Kafka Load

AI-READY

The Next-Generation Data Infrastructure for the AI Era

Agentic Analytics

Empowers autonomous AI agents to interact directly with your data through seamless MCP Server integration for real-time decision-making.

Hybrid Search

Provides the ultimate retrieval foundation for RAG by unifying Full-text search and Vector search within a single high-performance engine.

AI-Powered SQL

Embeds LLM capabilities directly within the database to enable SQL-driven large-scale text analysis, and semantic exploration.

AI Lakehouse

Leverages columnar JSON and multi-modal capabilities to master unstructured data at scale. It accelerates analytics, feature engineering, and data wrangling with seamless AI engine interoperability.

AI Observability

Seamlessly manage AI logs and traces with native integrations for Langfuse and the broader AI ecosystem. Gain full visibility into your AI stack with an ultra-low-cost observability solution.

UNIFIED

The Unified Engine Replacing ClickHouse, Trino, and Elasticsearch

vs ClickHouse

VeloDB/Doris outperforms ClickHouse through its advanced MPP architecture with a Cost-Based Optimizer (CBO) for efficient distributed joins and a strongly consistent primary key model for high-frequency real-time updates.

Execution time for TPC-H SF100 queries

Lower is better

vs Trino

VeloDB/Doris outperforms Trino in Lakehouse analytics by delivering 3x faster query speeds through its C++ vectorized engine and native local caching.

Execution time for TPC-DS 1TB queries

Lower is better

vs Elasticsearch

VeloDB/Doris delivers 10x higher cost-efficiency and 5x faster write throughput than Elasticsearch by leveraging columnar storage with ZSTD compression, inverted indexes for flexible text search, and a native MPP engine for complex SQL-based analytics.

HTTP Logs Benchmark

247 millions records in 32 GB, with 11 Queries

USE CASES

Built for Modern Data and AI Workloads

Customer-Facing Analytics

Purpose-built and architected for data applications to handle high-concurrency and low-latency analytics.

Explore Customer-Facing Analytics

Data Warehouse

Sub-second analytics on petabyte-scale lakehouse data. Open formats, unified workloads, no vendor lock-in.

Explore Data Warehouse

Observability

Streamline logs, metrics, and traces with a SQL-based observability stack, delivering lightning-fast insights at a fraction of the cost.

Explore Observability

Context Engineering

Store vectors, full-text, structured data, and JSON in one database. Serve fresh context to LLMs and agents in milliseconds.

Explore Context Engineering

User Stories

Observability

Talkie's Creator Migrates from Loki, Builds PB-scale Logging with Apache Doris.

Responds to keyword searches and aggregation queries from 1 billion log records within 2 seconds.
Deliver a log write throughput of 10 GB/s, while maintaining data latency within seconds.
The data compression ratio reaches 5:1 and tiered storage further reduces storage costs by 70%.

Read Now

Real-Time Analytics

JD Search Box's Real-Time Analytics Evolution with Apache Doris

Handles 10 billion rows of logs daily.
Achieves high throughput with 10,000 QPS and a minimum query latency of 150ms.
Enables real-time data ingestion by 60 million rows per minute.

Read Now

Alternative to Elasticsearch

Saving 80% costs by migrating from Elasticsearch to Apache Doris

Storage usage has dropped by 72%, while the overall operational costs have been cut by up to 80%.
Improves writing performance by 4x compared to Elasticsearch.
A unified Doris-based data platform, eliminating data inconsistency issues.

Read Now

Observability

Talkie's Creator Migrates from Loki, Builds PB-scale Logging with Apache Doris.

Responds to keyword searches and aggregation queries from 1 billion log records within 2 seconds.
Deliver a log write throughput of 10 GB/s, while maintaining data latency within seconds.
The data compression ratio reaches 5:1 and tiered storage further reduces storage costs by 70%.

Read Now

Real-Time Analytics

JD Search Box's Real-Time Analytics Evolution with Apache Doris

Handles 10 billion rows of logs daily.
Achieves high throughput with 10,000 QPS and a minimum query latency of 150ms.
Enables real-time data ingestion by 60 million rows per minute.

Read Now

Alternative to Elasticsearch

Saving 80% costs by migrating from Elasticsearch to Apache Doris

Storage usage has dropped by 72%, while the overall operational costs have been cut by up to 80%.
Improves writing performance by 4x compared to Elasticsearch.
A unified Doris-based data platform, eliminating data inconsistency issues.

Read Now

Observability

Talkie's Creator Migrates from Loki, Builds PB-scale Logging with Apache Doris.

Responds to keyword searches and aggregation queries from 1 billion log records within 2 seconds.
Deliver a log write throughput of 10 GB/s, while maintaining data latency within seconds.
The data compression ratio reaches 5:1 and tiered storage further reduces storage costs by 70%.

Read Now

Real-Time Analytics

JD Search Box's Real-Time Analytics Evolution with Apache Doris

Handles 10 billion rows of logs daily.
Achieves high throughput with 10,000 QPS and a minimum query latency of 150ms.
Enables real-time data ingestion by 60 million rows per minute.

Read Now

Alternative to Elasticsearch

Saving 80% costs by migrating from Elasticsearch to Apache Doris

Storage usage has dropped by 72%, while the overall operational costs have been cut by up to 80%.
Improves writing performance by 4x compared to Elasticsearch.
A unified Doris-based data platform, eliminating data inconsistency issues.

Read Now

INTEGRATION

Real-Time Analytics & Search DB

The Fastest End-to-End Engine from Ingestion to Insight

The Next-Generation Data Infrastructure for the AI Era

The Unified Engine Replacing ClickHouse, Trino, and Elasticsearch

Built for Modern Data and AI Workloads

User Stories

Seamlessly Integrate with Your Stack

Recent News