Selected work: connecting systems, data, and reporting

Funder and government reports took about a week to turn around, and quarterly reporting took two weeks to prepare.

Migrated reporting from Tableau to Power BI, redesigned data models with executive and BI leadership, and defined Dynamics 365 KPIs and reporting criteria for a federal funding program.

Reporting moved from a one-week turnaround to on-demand access. Quarterly reporting preparation dropped from two weeks to 3–4 days.

Qualifying incoming leads required significant staff time.

Led a five-phase initiative delivering a four-agent application built with Claude's API and structured Markdown instructions, coordinating stakeholders and technical delivery.

Lead qualification time fell by approximately 75%. This professional delivery example demonstrates workflow automation, not a new standalone service.

Technical delivery of 200+ enterprise-system enhancements.

Owned requirements, solution decisions, development, testing, and releases.

Shortened delivery cycles from 2–3 weeks to one week.

Internal requests were scattered across email, chats, spreadsheets, and informal follow-ups, making ownership, status, priority, comments, and reporting difficult to manage.

Designed a normalized Azure SQL backend with request, category, status, comment, and assignment tables, plus keys, constraints, timestamp logic, seed data, and a reporting-ready view.

Structured tracking, cleaner data entry, and a foundation for future Power Apps integration and BI reporting. Presented as a backend design, not a deployed client solution.

Scheduled BI refreshes depended on a gateway installed on a local workstation, creating a single point of failure when the machine slept, rebooted, disconnected, or became unavailable.

Designed a migration to a cloud-hosted Windows VM running the Power BI gateway while keeping the existing source, ingestion, storage, preparation, and query layers unchanged.

A refresh path with less workstation dependency, with planned monitoring, credential handling, backups, and future high-availability options. This is a migration plan, not a deployed client result.

Business users were spending too much time manually handling website intake, CRM data capture, record creation, and follow-up, slowing response times and making lead activity harder to track.

Designed a website-to-CRM intake workflow using a Real-time Journeys form, Dataverse custom activity records, and interaction tracking for advisory follow-up and reporting visibility.

Reduced manual handling by approximately 95%, with faster client response, cleaner lead capture, fewer errors, and better follow-up visibility.

External business records needed to be searchable inside a model-driven app without overwriting operational CRM accounts, creating duplicates, or relying on repeated manual imports.

Designed a low-code integration using a Power Platform custom connector, Power Automate, a separate Dataverse table, and scheduled append-and-update logic driven by stable external record IDs.

A controlled way to review external profiles, detect changes, and keep operational CRM records separate, with a proposed path to Azure Functions when low-code limits are reached.

Website inquiries can get lost in inboxes, copied manually into spreadsheets, or tracked without a clean contact history, making follow-up slower and harder to measure.

Built the inquiry workflow behind this website: Vercel forwards form submissions to Azure Functions for server-side validation and hCaptcha verification, with Azure SQL contact and inquiry records and Microsoft 365 notifications.

Valid submissions create inquiry records, contacts are matched by email, and message history is preserved. Future extensions could connect CRM, Teams, automation, or BI; these are not presented as already deployed.

Standard business rules were not enough to control the form behavior users needed, leaving room for missing context, inconsistent records, and manual field updates.

Implemented JavaScript-based model-driven form logic using onLoad behavior to validate values and auto-populate fields where configuration alone was not sufficient.

Improved data completeness and reduced entry errors through validation and automatic field population at the point of entry.

The existing Skyvia ingestion approach hit platform limits, creating scaling pressure for larger data movement and reporting needs.

Architected a Dataverse-to-Azure SQL pipeline using Azure Data Factory and Azure Blob Storage within the organization's existing cloud environment.

Processed 100K+ records per cycle, improved reliability, reduced manual work, created a cleaner reporting layer, and controlled cost through the existing Azure environment.