The database world has always forced developers into hard choices: prioritize transactional integrity, optimize for analytics, or accelerate search — but rarely all three at once.
Cantilux, a San Francisco–based startup founded in 2024, believes it has solved that tradeoff. The company has secured an $8 million seed round led by Lightspeed Venture Partners, with participation from Accel and several seasoned angels in the data-infrastructure ecosystem, to bring its vision to market.
Breaking the Database Tradeoff
Traditional architectures demand multiple systems: one database for transactions (OLTP), a warehouse or analytical engine for OLAP, and a separate layer for full-text search. That fragmentation creates lag, cost, and operational headaches.
Cantilux proposes a different path: a single, cloud-native engine capable of handling all three workloads in real time. By collapsing OLTP, OLAP, and search into one tier, the company aims to reduce latency, simplify data pipelines, and enable global-scale performance without manual sharding.
Why It Matters Now
Applications across fintech, gaming, and e-commerce increasingly depend on millisecond-level responsiveness and fresh data streams. Meanwhile, AI/ML pipelines demand immediate access to features for training and inference. Moving data between siloed systems slows everything down.
Cantilux positions itself as AI-native, designed to support low-latency queries that plug directly into machine learning workflows. The architecture also incorporates enterprise-grade security, with SOC 2 readiness, granular permissions, and zero-trust assumptions built in from the ground up.
What the Platform Offers
The company describes its product around four core pillars:
- Unified workloads: OLTP + OLAP + full-text search in a single system
- Elastic scalability: Instantly scales in the cloud without manual sharding
- AI-first design: Low-latency support for AI/ML pipelines
- Operational clarity: A sleek console with built-in deployment and monitoring tools
The developer experience was intentionally prioritized. Teams can spin up production-ready clusters quickly, with observability embedded into the console, rather than relying on external monitoring.
Early Traction and Results
Although it only emerged from stealth earlier this year, Cantilux already counts 20+ enterprise design partners. These early adopters report query speeds up to 10× faster than incumbent databases and note significant infrastructure savings from consolidating multiple systems into one.
One CTO from a global fintech partner summarized the appeal: eliminating the need to maintain separate transactional, analytical, and search stacks while still delivering sub-millisecond performance.
The Road Ahead
The fresh capital will be used to expand Cantilux’s engineering team, accelerate product development, and prepare for the launch of a fully managed cloud service later in 2025. The company is also in discussions with major cloud providers to make its engine available across developer ecosystems worldwide.
Snapshot
Funding: $8 million seed round
Investors: Lightspeed Venture Partners (lead), Accel, angel investors
Founded: 2024
Founder/CEO: Adrian Keller
Headquarters: San Francisco
Customers: 20+ enterprise design partners across fintech, gaming, e-commerce
Key differentiator: Unified engine for OLTP, OLAP, and full-text search in real time
Next milestone: Launch of managed cloud service later this year
Positioning
Cantilux’s thesis is straightforward: as data-driven applications grow more demanding, the split between transactional systems, analytical warehouses, and search engines is becoming unsustainable. The company is betting that enterprises will prefer a single database layer that eliminates compromises, lowers latency, and simplifies operations — especially as AI workloads intensify.
With fresh capital and early traction, Cantilux is positioning itself as the first true real-time, unified database platform for organizations that cannot afford to trade off speed, scale, or accuracy.
