Requests per second
Observed in measured CacheDB workloads.
CacheDB by houdata.
A transparent database cache for critical workloads.
CacheDB sits between your application and database, accelerating repeated reads while keeping invalidation controlled and visible.
- Transparent path
- Keep existing application logic.
- Targeted invalidation
- Refresh only what changes.
- Operator visibility
- Inspect the cache in real time.
Console preview. Video coming soon.
Works with critical data systems
- PostgreSQL
- MySQL
- IBM DB2
- Oracle
- SQL Server
Measurable improvement
Make repeated reads faster. Keep data changes controlled.
CacheDB is designed for teams that need a credible performance path without a rewrite of the systems they already trust.
Fewer database requests
For cacheable, read-heavy traffic with an appropriate policy.
Policy built from traffic
Analyze proxy signals and propose query-specific cache actions for review.
See the workload
Inspect query patterns, QPS, cache outcomes, latency and invalidations.
Results depend on workload, database, cacheability, policy, and infrastructure. Request the evaluation methodology.
Architecture
One intelligent layer between systems.
CacheDB answers repeated reads from Redis when safe, routes misses to the source of truth, and makes the resulting behavior observable.
Explore the technical architecture
Traffic-aware cache policy
Turn proxy traffic into a better Rhai policy.
CacheDB combines programmable hooks with an AI assistant that studies the operational context already visible at the proxy.
Inspect the traffic
Analyze cache metrics, query patterns, invalidation relationships and runtime signals to find inefficient behavior.
Propose the change
Generate a focused Rhai candidate for TTLs, cache eligibility, or selective invalidation instead of a generic recommendation.
Review before rollout
Validate candidates with the production engine without deploying them. Operators remain responsible for review and activation.
Enterprise-ready operations
Built for the controls around a critical database path.
CacheDB makes performance policy an operational system: observable, attributable, integrated with identity, and designed for resilient deployments.
Durable alerts
Evaluate and retain operational alerts, then deliver them through configured SMTP or webhook channels.
Auditable change
Record protected actions and policy deployments in an append-only audit history with export support.
High availability
Run multiple proxy instances and use Redis Sentinel so failover is discovered without proxy restarts.
Single sign-on
Use standards-based OIDC, group-to-role mappings, and a controlled emergency-access path for the console.
Evidence for operators
Understand the query. Review the policy.
CacheDB exposes the operational context needed to review cache outcomes and tune explicit policy with confidence.
A controlled evaluation
Observe. Decide. Act. Measure.
Start with the workload you already operate. Validate the result with evidence before broadening the deployment.
- 01
Observe
Identify repeated reads, load patterns and current constraints.
- 02
Decide
Model cache policy, TTLs and invalidation boundaries.
- 03
Act
Introduce CacheDB into a controlled path without rewriting application logic.
- 04
Measure
Compare latency, hit ratio and database load against the baseline.
Technical documentation
Documentation that stays close to the system.
Technical conversation
Evaluate CacheDB against your environment.
Tell us about the database, workload and performance constraint you are working with. We will start with a focused, evidence-led discussion.