Hi, I'm Luis Felipe Navarro.

Business Problem

The company was building a new underwriting platform to create a unified system across regions and lines of business. During the migration from the legacy application to the new platform, stakeholders needed a reliable way to detect data accuracy issues, misleading information and recurring migration errors.

Data Used

The analysis was based on underwriting data extracted from both the legacy application and the new platform. The objective was to compare migrated records, identify inconsistencies and highlight contracts requiring review or correction.

Process

We first worked with underwriters to understand where they were experiencing the most frequent data accuracy issues. After identifying the main error types, I designed SQL queries to extract the required information, apply business logic and validate the accuracy of migrated data.

The final output was a Power BI dashboard that allowed stakeholders to monitor migration quality by line of business, error type and contract. Users could review migration accuracy percentages and quickly identify contracts affected by issues such as pricing inconsistencies or manual input errors.

Dashboard Capabilities

• Monitor migration accuracy by line of business.
• Identify common data quality issues by category.
• Drill down from summary KPIs to individual contracts.
• Provide direct access to contract-level information for correction.
• Support underwriters and stakeholders with a clear exception report.

Impact

This solution significantly reduced the time required to identify migration errors and created an intuitive way for users to detect, understand and correct data quality issues during the migration process.

Detailed case study — soon available

WITH migration_check AS (
    SELECT
        c.line_of_business,
        c.contract_number,
        c.underwriting_year,
        c.pricing_amount AS legacy_pricing,
        n.pricing_amount AS new_platform_pricing,
        c.status AS legacy_status,
        n.status AS new_platform_status,
        c.last_update_date,
        CASE
            WHEN n.contract_number IS NULL THEN 'Missing in New Platform'
            WHEN c.pricing_amount <> n.pricing_amount THEN 'Pricing Mismatch'
            WHEN c.status <> n.status THEN 'Status Mismatch'
            ELSE 'Valid Migration'
        END AS migration_status
    FROM legacy_contracts c
    LEFT JOIN new_platform_contracts n
      ON c.contract_number = n.contract_number
     AND c.underwriting_year = n.underwriting_year
),
quality_summary AS (
    SELECT
        line_of_business,
        migration_status,
        COUNT(*) AS total_contracts,
        ROUND(
            COUNT(*) * 100.0 /
            SUM(COUNT(*)) OVER (PARTITION BY line_of_business),
            2
        ) AS status_percentage,
        COUNT(CASE WHEN migration_status <> 'Valid Migration' THEN 1 END) AS contracts_with_issues
    FROM migration_check
    GROUP BY line_of_business, migration_status
),
ranked_issues AS (
    SELECT
        line_of_business,
        migration_status,
        total_contracts,
        status_percentage,
        RANK() OVER (
            PARTITION BY line_of_business
            ORDER BY total_contracts DESC
        ) AS issue_rank
    FROM quality_summary
)
SELECT
    line_of_business,
    migration_status,
    total_contracts,
    status_percentage,
    issue_rank
FROM ranked_issues
WHERE migration_status <> 'Valid Migration'
ORDER BY line_of_business, issue_rank;

Business Problem

While managing several applications at the same time, the team needed an effective way to measure application support performance from a user-oriented perspective. The goal was to track SLA performance, user tickets, issue types, response times and the overall quality of application maintenance.

Data Used

The dashboard used data from ServiceNow and Azure DevOps, including user tickets, issue categories, application references, response times, resolution status and support activity.

Process

We began by defining the KPIs needed to evaluate application maintenance. After identifying the right KPIs, I selected and structured the available data, created Power BI measures and designed visuals that allowed the team to follow application performance in a simple and consistent way.

Dashboard Capabilities

• Track SLA performance across different applications.
• Monitor the number of tickets raised by users.
• Classify tickets by type, from information requests to critical bugs.
• Measure response and resolution times.
• Provide management with a clear view of application support quality.

Impact

The dashboard made application follow-up easier and more intuitive for the team. It also provided a clear and concise way to explain application maintenance performance to management and support operational decisions.

Final Output

The final solution was an interactive Power BI dashboard that centralized key maintenance indicators and helped the team monitor application performance, ticket management and SLA compliance.

Detailed case study — soon available

Business Problem

The team needed to measure the performance of a complex internal application. The main challenge was translating highly technical transaction processing data into simple, readable and business-oriented KPIs that could be understood by product teams and management.

Data Used

The analysis used transaction, contract, message and accounting-related data extracted from Databricks. SQL queries were developed to structure the data and create the foundation for Power BI measures and visuals.

Process

We started by defining the KPIs with the Head of Application. Then, I reviewed the available data in Databricks, identified missing fields and developed SQL logic to extract the information needed for the dashboard.

After preparing the dataset, I built Power BI measures and designed the dashboard to present application performance, transaction origin, message status and business activity in a clear and explainable way.

Dashboard Capabilities

• Monitor transaction processing performance over time.
• Classify transactions by origin, such as manual, batch or electronic processing.
• Track message status, business activity and contract-related information.
• Transform technical application data into readable KPIs.
• Support product and management teams with clear performance insights.

Key Insight

Collaboration between the Data Analyst and the Head of Application was essential. Defining the right KPIs and agreeing on how to display them helped transform a complex operational process into an intuitive monitoring solution.

Final Output

The final dashboard helped monitor application performance, identify operational patterns and communicate complex transaction behavior in a simple way.

Detailed case study — soon available

SELECT
    transaction_date,
    transaction_number,
    contract_number,
    transaction_code,
    CASE
        WHEN message_id IS NOT NULL THEN 'Electronic Processing'
        WHEN update_user IN ('AUTO', 'BATCH') THEN 'Batch'
        ELSE 'Manual'
    END AS transaction_origin,
    message_status,
    business_activity,
    original_currency_amount
FROM accounting_transactions
WHERE transaction_date >= DATE('2024-01-01');