I help enterprise organizations get real value from their data, and help teams turn that into working systems.
My job is to make complex things make sense. I help a business team understand what's possible with their data, and I help an engineering team understand what the business needs. I'm the person in the middle keeping everyone on the same page.
I've spent my career in technical operations, from field workforce management to the data foundations that let organizations take AI seriously. Not "AI" as a buzzword, but agentic capabilities that make people's work better.
Right now I'm focused on how fiber infrastructure companies can use their operational data more intelligently: how technicians close out jobs, how managers spot patterns they'd miss by hand.
Outside of work I'm a dad, a casual golfer, and someone who builds side projects to see what's possible. This website is one of them, built as a way to learn and to show something beyond a resume bullet point.
Served as product owner across customer service provisioning, dispatch and scheduling, field operations tooling, and reporting and analytics within a greenfield BSS/OSS build, coordinating 100+ engineers and contributors across development teams to deliver the modules that form the operational foundation the business runs on today. Translated business needs into technical scope and kept order management and provisioning data flowing cleanly between systems like ServiceNow, Salesforce, and Snowflake.
Preserved the full production instance of a sunsetting application by designing a tiered Snowflake archive: a Raw schema of 44 tables for the source data, a Curated schema of 19 tables for the reports managers rely on day to day, and a Stage schema wired to Azure Blob to hold 90GB of attachments. Built the cross-cloud integration and configured the blob permissions for secure read access, then created a staged view with pre-signed URLs so anyone could search attachments by name, type, or size and open them through time-limited links. Set up dedicated user roles and a separate warehouse to keep access controlled and compute costs in check.
Returned roughly 18 hours a week to operations teams by finding manual work and eliminating it. An address validation and approval pipeline routes requests from a shared inbox into a tracked SharePoint workflow with automated status and approval alerts, handling about 50 requests weekly and saving around 12 hours. An end-to-end Snowflake solution generates, distributes, and archives eight market-specific reports, replacing the manual filtering and spreadsheet prep behind each one and saving about 6 hours weekly, or 312 hours a year. A third pipeline feeds competitive market news into a Power BI dashboard that informs market expansion decisions.
Designed a tiered data strategy for a regional fiber infrastructure provider and presented it to the CTIO and Director, giving leadership a shared language for where the organization sits versus where it needs to be. The framework moves from clean, reliable source data (bronze) to integrated analytics (silver) to AI-ready, agentic pipelines (gold), and sets the sequencing for what gets built next.
Built this site collaboratively with Claude AI, using it as both a learning exercise and a real artifact to point to. The whole thing was a back-and-forth: I provided context and direction, Claude helped with design and code, and I shaped it into something that felt like me.
Built and now running an autonomous knowledge system on top of Obsidian, using Claude Code as the engine that reads, writes, and connects notes across a PARA-structured vault. Added two content agents on top of the base system: one that pulls and summarizes YouTube transcripts, another that researches topics live via web search and scraping, both filing structured, source-linked notes back into the vault automatically. The last piece was a scheduled agent, a Task Scheduler job that runs unattended against a watchlist on its own scoped permissions, researches what's due, and logs every run for review. At its current research cadence, that agent alone is saving an estimated ~2 hours a week versus doing the same searching, reading, and synthesis by hand.
Researched and developed a strategy for an AI-powered assistant for field technicians, designed to provide real-time guidance on job code selection during installs and service calls. The same system would surface behavioral insights from closeout notes to help managers spot trends they'd miss by hand. Currently working through the data foundation requirements before moving to implementation.
Working through what it takes for an enterprise field operations company to go from "we have data" to "we have AI that does something useful." Covers data quality, tooling, team readiness, and the organizational change that nobody wants to talk about. Currently evaluating Snowflake Intelligence, Cortex Analyst, and the Snowflake semantic layer as core components of the implementation path.
One framework I keep coming back to: organizations don't jump straight to AI. They earn it. You build the foundation first, or everything on top of it wobbles.
I use a bronze-silver-gold model to help teams see where they stand versus where they want to be. It sounds simple, but getting leadership to admit they're at bronze is hard when everyone assumed they were already at gold.
Once you're honest about that, the path forward gets clearer.
Consistent, clean, reliable data at the source. Getting systems to agree on basic facts before anything else.
Connected data across systems. Dashboards that tell a story. Teams that use the reports being built for them.
AI that acts on data in real time. Predictive signals. Systems that help people do their jobs better without getting in the way.