1
Data Catalog
Select the Data Source

Browse the role-filtered Data Catalog and select the source table. Row counts, descriptions, and sensitivity flags are visible before any query is written.

Actual Data Source Selected
DATAWAREHOUSE.KINETIQ_INSIGHTS.HEALTHCARE_DATA
2
Columns
Choose the Columns in Scope

Select exactly which columns the AI will work with. Only schema metadata is used — no raw data is ever sent to the AI.

Actual Columns Selected
PAYER_NAME HCPCS OUTCOME DATE_OF_SERVICE THERAPIST_NAME
3
Use Case
Define the Business Question

Describe what you need in plain English. The AI validates the request for clarity and SQL suitability.

Actual use case entered
"Create a comprehensive health plan performance report analyzing medical assessment outcomes across all available historical data from DATAWAREHOUSE.KINETIQ_INSIGHTS.HEATHCARE_DATA. Build a pivot table with PAYER_NAME as rows and four outcome columns (Recommended, Not Recommended, Not Recommended Downgraded, Not Recommended Upgraded), showing both absolute count and percentage distribution for each outcome per payer. Filter out null or empty OUTCOME values, calculate row-level percentages (outcome count / payer total * 100), and sort alphabetically by PAYER_NAME. Include frozen headers with alternating opacity styling and responsive layout. Possible outcome values are Recommended, Not Recommended, Recommended Upgrade, Recommended Downgrade."
4
SQL Generation
AI Generates & Tests the Query

The AI produces a validated SELECT query tailored to the use case. The query is tested against live data before proceeding — only SELECT statements are permitted.

5
Data Quality ✓ Approved
Validate & Approve the Data

Query results are previewed before any visualization is created. A named approver signs off on data quality — this gate is required and tracked.

Actual Approval Recorded
✓ Approved by Ben Marchbanks  ·  2026-01-03
6
Refine & Release
AI Generates the Visualization — Then You Refine It

The AI recommends the best visualization type and generates the raw program code required for an interactive rendering. Each refinement is a natural language instruction — no code editing required. Every iteration is stored and reusable.

Actual Refinements Applied
Style "Use contrasting text colors for elements with dark background colors."
▶ Data Dictionary
Column Definition Domain Examples Synonyms
Not_Recommended_Count The total number of medical assessments where the outcome was 'Not Recommended' for a specific payer. This represents cases where the medical review determined that the proposed treatment, service, or intervention should not proceed as originally requested. Healthcare 78, 156, 23, 445, 89 Denied Count, Rejected Assessments, Unfavorable Decisions, Declined Cases, Negative Outcomes
Not_Recommended_Downgraded_Count The number of assessments where the original recommendation was downgraded during review, resulting in a 'Not Recommended' final outcome. These cases initially had a more favorable assessment that was later reduced to a negative recommendation through the review process. Healthcare 15, 42, 7, 89, 26 Downgraded Denials, Reduced Recommendations, Lowered Assessments, Degraded Outcomes, Diminished Decisions
Not_Recommended_Downgraded_Pct The percentage of total assessments that were downgraded from a higher recommendation to 'Not Recommended' status. This shows the rate at which initially favorable assessments were reduced to negative outcomes during the review process for each payer. Healthcare 8.3, 15.7, 4.2, 22.1, 11.9 Downgrade Rate, Reduction Percentage, Degradation Rate, Lowered Decision Rate, Diminished Outcome Rate
Not_Recommended_Pct The percentage of total assessments for a payer that resulted in a 'Not Recommended' outcome. This metric shows the denial or rejection rate for each payer, calculated by dividing not recommended cases by total assessments and multiplying by 100. Healthcare 32.5, 76.2, 10.8, 54.4, 21.1 Denial Rate, Rejection Percentage, Unfavorable Rate, Decline Rate, Negative Outcome Rate
Not_Recommended_Upgraded_Count The number of assessments where the recommendation was upgraded during review but still resulted in a 'Not Recommended' final outcome. These represent cases that received some improvement in assessment but remained below the threshold for approval. Healthcare 8, 34, 12, 67, 19 Upgraded Denials, Enhanced Rejections, Improved Non-Approvals, Elevated Declines, Boosted Negative Outcomes
Not_Recommended_Upgraded_Pct The percentage of total assessments that were upgraded during review but still concluded with 'Not Recommended' outcomes. This metric indicates the rate of cases that received favorable adjustments but remained insufficient for final approval by each payer. Healthcare 3.1, 12.4, 5.8, 18.9, 7.6 Upgraded Denial Rate, Enhanced Rejection Rate, Improved Non-Approval Rate, Elevated Decline Rate, Boosted Negative Rate
PAYER_NAME The name of the health insurance company, health plan, or organization that provides coverage and payment for healthcare services. This represents the entity responsible for reviewing and making coverage decisions on medical assessments and treatments for their members. Healthcare Blue Cross Blue Shield, Aetna Health Plans, UnitedHealthcare, Cigna Corporation, Humana Inc Insurance Provider, Health Plan Name, Coverage Provider, Insurer Name, Plan Sponsor
Recommended_Count The total number of medical assessments where the outcome was determined to be 'Recommended' for a specific payer. This represents cases where the medical review process concluded that the proposed treatment, service, or intervention should proceed as originally requested. Healthcare 245, 89, 567, 12, 334 Approved Count, Favorable Decisions, Positive Outcomes, Endorsed Cases, Accepted Assessments
Recommended_Pct The percentage of total assessments for a payer that resulted in a 'Recommended' outcome. This is calculated by dividing the recommended count by the total assessments and multiplying by 100, showing the approval rate for each payer. Healthcare 67.5, 23.8, 89.2, 45.6, 78.9 Approval Rate, Recommendation Percentage, Positive Outcome Rate, Endorsement Rate, Favorable Decision Rate
Total_Assessments The complete count of all medical assessments processed for a specific payer, regardless of outcome. This represents the total volume of cases reviewed and serves as the denominator for calculating all percentage metrics for that payer's performance analysis. Healthcare 1,250, 896, 2,145, 567, 3,421 Total Cases, Assessment Volume, Total Reviews, Case Count, Total Evaluations