I build the systems that decide what a business does next — metric layers, experiments, and the AI agents that put insight directly in a leader's hands — then help Product and Engineering ship it.
Sixteen years in analytics and AI, including nine-plus at Walmart, spent in the seam between what the data says, what engineering can build, and what the customer actually needs. I currently lead a 25+ person organization across four pillars — Self-Serve/Chatbots/AI, CRM, Workforce Management, and Ops — and have shipped alongside Product teams in customer service, fulfillment, facilities, fuel & convenience, and store innovation.
My pattern is consistent: move analytics from reporting what happened to shaping what the business does next. That means building the metric repositories and governance that make enterprise data trustworthy, the experiments that prove what actually works, and increasingly, the AI agents that put certified answers in a leader's hands without a request queue in between.
The work overlaps constantly — an AI launch is also a product bet, a data platform is also a business-strategy call. These buckets are just the four lenses people usually ask about.
Agentic AI, GenAI, computer vision, and conversational AI — built to change a decision, not just add a feature. My focus is always the measurement framework underneath: how do we know it worked?
Built an agentic AI insights product that serves as the org's governed metric repository. Leaders query certified definitions in natural language and get narrative answers with context — not just tables — and the semantic layer keeps the number the same no matter who asks. It absorbed the recurring ad hoc analysis load and now powers “4-in-the-box” product, ops, and engineering reviews.
Owned the measurement framework — containment, resolution, cost per contact — that leadership used to judge the shift from traditional IVR to an AI voice assistant. Self-serve rate rose from 30% to 70%, with CSAT climbing alongside it rather than trading off against it.
Partnered directly with Product on AI-based voice and chat agents on the CRM platform — requirements and problem framing, pilot design, launch readiness, adoption tracking, and post-launch iteration. Built sentiment analysis into the measurement layer to track how customers actually felt about AI-handled interactions turn by turn, not just whether the contact resolved — catching tone and satisfaction drift early enough to fix it before it showed up in CSAT. Part of a broader CRM AI push that delivered $36M in realized savings.
Directed the data analytics and ML team behind a computer-vision shrink-prevention platform. Behavior and shrink-pattern analysis drove material product improvements, preventing $330M in loss and producing a filed patent (Virtual Cart Optimization Tool).
I build the analytics that decide what gets built — instrumentation, funnels, experiment design, and the success metrics a product is judged on — embedded with Product from problem definition through launch and scale.
Own product analytics for the full service portfolio — instrumentation, journey funnels, containment and escalation analysis, and self-serve deep dives that tell Product exactly where customers dropped off and which experiences to rebuild first.
Partnered with the Walmart+ product team to launch a member fuel benefit, building the adoption and redemption tracking that connected fuel behavior to membership growth and retention from the first day of launch.
Designed and built Fuel & Convenience's first automated pricing tool in 2020: continuous WTI/Brent crude and Kalibrate competitor-pricing feeds, a configurable radius and price-gap threshold, and a BI layer that quantified the dollar cost of any unaddressed breach before alerting the pricing team via Power BI and mobile. Replacing manual, once-a-day price checks with 24/7 monitoring drove an 8.5% quarter-over-quarter increase in gallons sold.
Partnered with Product to deploy live refrigeration and HVAC telemetry monitoring, giving field teams a signal before equipment failed — turning a maintenance queue into a proactive product experience.
Established definitions, architecture review, and sign-off processes that gave Product one trusted source of truth for roadmap and prioritization — replacing dashboard sprawl with an actual decision system.
I work at the seam where physical footprint, operating model, cost structure, and customer experience meet — framing the right problems and converting complex operational challenges into scalable capabilities.
Built the strategy and data foundation for a new division from zero, standing up the metrics an entire business needed to operate — and drove a 26% merchandise sales lift by connecting fuel traffic to inside-store purchases.
Helped shape the build-out and operating case for Internal Consolidation Centers, modeling flow, capacity, and cost-to-serve to inform how volume moved through the network — part of a 35% GMV lift for the fulfillment business.
Translated network economics — cost to serve, capacity, carrier performance — into a one-page format used every week across Weekly Business Reviews, making complex trade-offs legible fast for a multibillion-dollar business.
Built the data and IoT foundation for Facilities Maintenance — architecture, telemetry flows, asset hierarchy — shifting the operating model from ticket response to prevention and informing capital versus expense decisions on repair-or-replace.
Enterprise data only creates value if people trust it and can actually get to it. I build the metric repositories, governance, and self-service layers that make that true at scale.
Owned capacity, row-level security, and self-service enablement for an enterprise analytics platform serving 300,000+ users, and designed a data-lake landscape that shifted the org toward self-service.
Shipped natural-language self-service so senior leaders could query their own data in free-form text — then trained them on it, treating adoption as seriously as the technical build.
Ran control-group testing and IVR modernization measurement, built AHT trackers, and recovered IEX forecasting accuracy from red to green — turning experimentation into a standing capability rather than a special project.
Used text analytics to diagnose root causes across the bottom 10% of Sam's Club locations and target intervention precisely — grounded by running field inspections as first-hand user research.
Built a weighted, six-factor model that scores every contact Low, Medium, or High complexity from contact reason, workflow count, talk %, repeat calls, transfers, and genuine/non-genuine intent — giving CES a defensible, auditable score to strategize call routing, headcount and skills planning, and scheduling around the real complexity mix instead of raw contact volume.
Live, interactive artifacts — click through to explore the real thing. Hover any card to pause and take a closer look.
Frameworks and points of view, written up so the thinking travels beyond one meeting.
Click through to verify directly with the issuing institution. University credentials carry a Cornell banner; platform certifications carry the issuer's mark. Hover any card to pause and take a closer look.
Sourced from LinkedIn recommendations, shown as received. Hover any card to pause and take a closer look.