Kits
Reusable evaluation frameworks
A Kit bundles Bricks into a complete rubric, scorecard, risk screen, or workflow-specific evaluation. Run one Kit and get the same structured fields back across every conversation in that workflow.
Discovery Signals Kit
One framework, many typed checks.
Definition
A Kit groups checks for one job
Bricks answer individual questions. Kits decide which questions belong together for a specific workflow such as discovery, QA, MEDDICC, renewal risk, or AI agent evaluation.
Composed from Bricks
Choose the checks that matter for a motion and keep each result separate instead of flattening everything into one score.
Mapped to a rubric
Model MEDDICC, BANT, discovery quality, support QA, compliance, or your own internal process as a reusable Kit.
Run as one unit
Pass a Kit code to the API and Semarize runs every Brick in the Kit against the conversation content.
Versioned over time
Change a Kit deliberately without corrupting historical analytics. Runs stay tied to the Kit version that produced them.
Reusable by teams
Use the same Kit across reps, regions, products, or support queues when the same standard should apply.
Ready for audit
A Kit produces structured fields, confidence, reasons, evidence, and version identifiers that downstream systems can store.
How it works
Assemble once. Run repeatedly.
Kits give teams a stable contract for a conversation workflow. The same questions are asked in the same shape every time.
Choose the motion
Start with the business workflow you need to evaluate, such as discovery, renewal risk, QA coverage, or data enrichment.
Select the Bricks
Add the checks that define the framework. Each Brick keeps its own type, rubric, evidence, and result.
Attach context
When a framework depends on product, pricing, policy, or playbook context, attach a knowledge base to the Kit.
Run the Kit
Send a transcript plus Kit code. The response returns every Brick result under the same Kit version.
Examples
Kits for real operating work
A Kit should match the decision or workflow that needs structured conversation data, not a generic dashboard category.
Discovery Quality Kit
Measures whether the rep uncovered pain, stakeholders, timeline, budget, success criteria, and a committed next step.
Forecast Risk Kit
Flags deal risk from missing budget, vague timeline, legal blockers, procurement friction, and competitor pressure.
MEDDICC Kit
Turns sales methodology adoption into discrete fields that can be trended across reps, teams, segments, and quarters.
Support QA Kit
Evaluates support conversations against quality, compliance, empathy, resolution, and escalation criteria.
CRM Enrichment Kit
Extracts the conversation facts that should update CRM fields without relying on memory or manual note entry.
AI Agent Evaluation Kit
Checks whether an agent or workflow handled a conversation according to approved rules and expected outcomes.
Comparison
Kits are not static scorecards
A scorecard is usually a manual review form. A Kit is a repeatable framework for systems that need comparable data.
Static scorecard
Semarize Kit
Operations
Frameworks that survive change
Teams need to improve their rubrics without losing trust in historical data. Kit and Brick versions make that easier to manage.
Immutable versions
Runs record which Kit version produced the result, so old analytics remain explainable after the framework changes.
Comparable fields
Because every run returns the same fields, teams can trend behaviour by rep, segment, account type, or quarter.
Downstream routing
Use Kit output to update records, alert managers, queue QA review, enrich a warehouse, or trigger a workflow.
Ready to build
Assemble your Kit. Run it with one API call.
Start with the framework your team already uses, then convert each criterion into a Brick inside a Kit.
No card required. Start testing in minutes.
Questions
Frequently asked questions
Short answers about how Kits work, how they relate to Bricks, and where they sit in the Semarize API workflow.
From the blog
Further reading
Practical pieces on turning Bricks into reusable frameworks for scoring, MEDDICC, QA, and repeatable conversation evaluation.