██╗ █████╗ ██╗ ██╗███████╗ ██████╗ ██████╗ ███████╗
██║ ██╔══██╗██║ ██╔╝██╔════╝██╔═══██╗██╔══██╗██╔════╝
██║ ███████║█████╔╝ █████╗ ██║ ██║██████╔╝███████╗
██║ ██╔══██║██╔═██╗ ██╔══╝ ██║ ██║██╔═══╝ ╚════██║
███████╗██║ ██║██║ ██╗███████╗╚██████╔╝██║ ███████║
╚══════╝╚═╝ ╚═╝╚═╝ ╚═╝╚══════╝ ╚═════╝ ╚═╝ ╚══════╝
The intelligent control plane for Apache Iceberg — autonomous compaction, maintenance, multi-engine routing, observability, and agentic AI. Connect in minutes. No infra changes. No vendor lock-in.
Real benchmarks from production-grade Iceberg tables across multiple engines and cloud providers.
| 80% | 12× | 95% | 100% |
|---|---|---|---|
| lower CPU & storage costs | faster queries after optimization | faster compaction vs. Apache Spark | maintenance automation |
| 29× faster vs. S3 Tables on 1 TB TPC-DS | set policies once — LakeOps runs the loop |
LakeOps connects to your existing catalogs, storage, and query engines — then continuously monitors cross-system telemetry and autonomously executes the right operations at the right time. Your data never moves.
Catalogs LakeOps Control Plane Engines
┌──────────────┐ ┌─────────────────────────────┐ ┌──────────────┐
│ Glue │ │ Observability & Insights │ │ Trino │
│ REST/Polaris │◄───────►│ Intelligent Compaction │◄───────►│ Spark/Flink │
│ Nessie │ │ Adaptive Maintenance │ │ Snowflake │
│ Gravitino │ │ Multi-Engine Query Routing │ │ Athena │
│ S3 Tables │ │ Policies & Governance │ │ DuckDB │
│ Lakekeeper │ │ Agentic AI & MCP │ │ StarRocks │
└──────┬───────┘ └──────────────┬──────────────┘ └──────┬───────┘
│ │ │
└────────────────────────────────┼────────────────────────────────┘
▼
┌─────────────────────┐
│ S3 · GCS · ADLS │
│ Iceberg tables │
└─────────────────────┘
Full Iceberg benefits. Snowflake-level ease. — without building Netflix-scale infrastructure.
|
Compaction — Rust engine on Apache DataFusion. Query-aware sorting by the columns your workloads actually filter on. Triggered by telemetry, not cron. Learn more → Maintenance — Adaptive loop that scores every table and sequences compaction → snapshot expiry → manifest rewrite → orphan cleanup. Learn more → Observability — Table health tiers, proactive insights, engine metrics, and a unified event history for every operation. Learn more → |
Query routing — One SQL endpoint across Trino, Spark, Snowflake, Athena, DuckDB, and Flink. Route by cost, latency, or throughput. Learn more → Agentic AI — MCP interface with schema discovery, wire-compatible PostgreSQL/MySQL/Arrow Flight, and layered guardrails for AI pipelines. Learn more → Governance — One-toggle policies for compaction, retention, orphan cleanup, and manifest optimization — catalog-wide or per-table. Learn more → |
| 1 Connect catalogs & engines |
→ | 2 LakeOps collects telemetry & scores table health |
→ | 3 Autonomous optimization runs continuously |
Works with AWS · Azure · GCP · Snowflake · Databricks · Delta Lake · Iceberg · Spark · Flink · StarRocks · Lakekeeper and more. SOC 2 Type II · SSO · RBAC · on-premise available.
→ Get started in 10 minutes · Explore the platform · See solutions
We build in the open and contribute to the Apache Iceberg ecosystem.
One query. Any engine.
Universal SQL query router and proxy written in Rust. Connect any SQL client over Trino HTTP, PostgreSQL wire, or MySQL wire — route to Trino, DuckDB, StarRocks, Athena, or ClickHouse with automatic dialect translation.
Client (Trino CLI / psql / mysql)
│ native protocol
▼
QueryFlux ── routing + dialect translation ──► Trino / DuckDB / StarRocks / Athena / ClickHouse
Pair with LakeOps for table-aware routing: as tables are compacted and sorted, more engines become viable for each query shape.
Rust · SQL · Trino · DuckDB · Arrow Flight SQL · Apache 2.0
→ queryflux.dev · Star on GitHub · Releases
| Resource | Description |
|---|---|
| Documentation | Step-by-step guides for every platform feature |
| Blog | Iceberg operations, migrations, and lakehouse architecture |
| Managed Iceberg in 2026 | The nine components every production data lake needs |
| Intelligent Lakehouse | Build like Netflix — without Netflix-scale engineering |
| Multi-Engine Routing | Route queries across engines on shared Iceberg tables |
| Trust Center | Security, compliance, and enterprise readiness |
lakeops.dev · Book a demo · Docs · QueryFlux
Apache Iceberg® is a registered trademark of the Apache Software Foundation.