Business-specific test cases
Giskard supports tests that describe what your assistant should and should not do in your domain, beyond generic quality metrics.
SERVICES / EVALUATION PLATFORMS
Functional quality testing with Giskard.
Independent planning and implementation for teams considering or using Giskard, focused on functional quality: contradictions, omissions, inappropriate refusals and business-rule failures.
Where Giskard fits
Knowledge assistants
Its open-source tooling includes a RAG evaluation toolkit for generating and scoring question sets against your knowledge base.
Collaboration with domain experts
Test cases can be reviewed and refined by non-engineers, which helps keep tests aligned with business reality.
Functional failures we target
Contradictions
Answers that conflict with your documents, your policies or the agent's own earlier statements in the conversation.
Omissions
Answers that are technically correct but leave out conditions, exceptions or next steps the user needed.
Inappropriate refusals
The agent declines legitimate requests it should handle, frustrating users and pushing work back to staff.
Out-of-scope answers
The agent answers questions it should redirect, or invents policy where none exists.
Retrieval gaps
Questions where the right document exists but wasn't retrieved, so the answer is incomplete or generic.
Rule violations
Eligibility, limits and required disclosures your business defines, expressed as testable checks.
How we help
Turn policies into tests
We translate business rules and expert knowledge into reviewed Giskard test cases with clear pass criteria.
Generate, then review
Generated question sets speed up coverage; expert review decides which cases are admitted to the benchmark.
Calibrate the judgment
Compare automated verdicts with reviewer labels, document disagreements and keep a held-out set for honest comparison.
Related services
Evaluation Assessment
A baseline for one Giskard agent or workflow, with prioritized findings and a release recommendation.
Explore serviceHallucination Testing
Measure how often your Giskard agent invents facts, policies or actions, and which fixes reduce it.
Explore serviceManaged Evaluations
Recurring tests as your Giskard agent, prompts, models and data change, with a monthly quality report.
Explore serviceFrom scope to handover
Start with your inputs
Representative tasks, knowledge sources, expected outcomes, platform access and a domain reviewer.
Keep reusable assets
A versioned test suite, scoring definitions, a baseline report, prioritized findings and a documented rerun process.
Support after handover
We can retest fixes, extend coverage to new Giskard workflows or move to recurring Managed Evaluations as your agent changes. Platform usage and licenses stay in your account.
Make your next decision with evidence.
Start with one workflow, clear success criteria, and an evaluation plan your team can use.