Automate always-on, performance-protected BigQuery optimization
Rabbit continuously tunes BigQuery cost and performance. Recommendations by default, automation when you opt in.
32%
BigQuery spend reduction
40%
performance improvement
400hrs+
engineering hours reclaimed monthly on average
Continuous tuning across every BigQuery cost lever
Optimizing one lever in isolation leaves savings on the table. Rabbit optimizes slots, pricing, storage, reservations, caching, and SQL as one system.
Max Slot Optimizer
Flatten spiky slot usage and cut slot consumption, without starving important workloads.
Rabbit assigns priority rules (1-5) per job and tunes max slots continuously so capacity matches real demand.
Flatten usage for higher commitment utilization
Reduce slot consumption when workloads compete
Protect latency-sensitive and high-priority jobs
47% spend ↴ at Nordstrom · 30% at Lufthansa
Job-level Pricing Model Optimizer
Route every job to the cheapest pricing model in near real time.
Rabbit evaluates on-demand vs capacity per job against live commitment utilization, including multi-statement scripts and jobs with no run history.
Near real-time per-job routing
Route across projects to avoid the 2k on-demand slot limit per project
Airflow & dbt plugins, or open-source BQ proxy (no pipeline code changes)
+15% savings at Nordstrom
Reservation Optimizer
Automate reservations, project assignments, and capacity prioritization from real job history without manual tweaking.
Rabbit creates and reconciles reservations so each workload lands on the right pricing model and rightsized capacity, with ramp profiles from scale-on-demand to cost-optimized.
Configurable priority profiles (on-demand · 1 min · 5 min · 10 min · 30 min)
Right pricing model on every workload, reservations stay aligned
Proactive slot quota management
32% cost saved at Lufthansa
Storage Model Optimizer
Pick logical vs physical storage billing per dataset, and apply your time-travel retention settings.
Rabbit analyzes access patterns across your estate and automates the cheaper billing model per dataset. Set time-travel retention limits and Rabbit applies them.
Automatic logical vs physical per dataset
Your time-travel retention policy, applied automatically
Savings estimate before you apply
25% lower storage costs
SQL Optimizations
Ship validated, automatic SQL rewrites with a savings estimate on each one.
Rabbit generates alternatives using BigQuery-specific knowledge, then tests equivalence before proposing a change.
Optimized SQL ready to apply
Savings estimate attached to each rewrite
Result-equivalence validation
16% average savings on query costs
Agentic optimization PRs
Turn existing Rabbit platform recommendations into ready-to-merge PRs.
Rabbit's Recommendation Applier automates the implementation step: it converts validated Rabbit recommendations into reviewable pull requests, while Build provides the developer workflow and context.
Applies existing Rabbit recommendations via automated PRs
Reviewable, ready-to-merge optimizations in Git
Full shift-left workflow on Build
Safe for enterprise BigQuery at scale
Rabbit acts autonomously on production BigQuery. You set the guardrails. For certifications, residency, and how Rabbit handles access, see Security
Performance protected
01
Recommendations first
02
SQL validated
03
For the teams that run BigQuery
Same platform. Different priorities. Pick the path that matches your role.
FAQ
