Built for Control.
Engineered for Action.

Walt integrates human decision-makers into every step that needs it — from metric approvals to SQL reviews. Your team can monitor agent behavior, review proposed changes, and enforce logic gates before anything goes live.
• Approve or reject agent-generated SQL
• Review and version metric definitions
• Add custom prompts, rules, or constraints
• Escalate high-risk queries to humans

Walt augments your team — not replaces them. Human oversight is built into the loop.
‍

Walt can ingest and adapt to your custom schemas, dbt models, LookML layers, and business metric logic — building its contextual layer on top of what you’ve already defined.
• Import or infer join paths and model logic
• Map business metric definitions to tables
• Tag sensitive fields or override assumptions
• Fully compatible with dbt, Looker, and custom stacks
‍
Walt meets you where you are — and builds intelligence around your existing work.

From onboarding your first schema to running Walt across multiple business units, our enterprise support team works with your leads to ensure success.
• Technical workshops with data & analytics teams
• Integration assistance with internal tools
• Custom use case onboarding
• Dedicated success partner & Slack channel

We don’t just hand you Walt. We build with you until your team is confident in using it.

Walt comes with a built-in evaluation framework designed for enterprise-grade quality assurance. Every answer is measured against goals like accuracy, consistency, cost, and latency — then improved continuously via agentic execution.
• 100+ test questions across common data scenarios
• Track performance by data source or metric
• Compare human vs. agent results
• Run evaluations on dev, staging, and production

Accuracy isn’t a promise — it’s a measurable output in Walt.