Bring me the business problem.
I'll own the technical work.

Challenge

Enterprise teams had large proprietary datasets, but getting an answer still meant working through spreadsheets or waiting on a fixed BI report. Every client had different sources, workflows, and questions. I also inherited six production repositories after the previous team left without documentation or a handoff.

What We Built

Automated pipelines pull each client's data into one system. Custom AI agents and natural language interfaces let teams query dashboards and complex datasets for real estate intelligence, marketing analytics, and operational decisions. The commute pipeline needed a different approach, so I replaced serial SQS polling with parallel Step Functions fan-out. Lambda memory use fell by 50% to 75%.

Challenge

The founders wanted Parscape to cover nearly the whole golf experience in one place. Golfers would be able to find where to play, practice, learn, shop, and be entertained, then earn rewards for using the platform. They needed a technical co-founder to build the web app, mobile app, backend, and infrastructure.

What We Built

The product spans a web app, React Native app, backend, and infrastructure. Headless Shopify powers the multi-brand marketplace for independent golf companies. Storyblok handles editorial content, while Algolia searches thousands of courses, simulators, learning centers, retailers, and other golf businesses. The platform also has user accounts, tiered cash-back rewards, subscriptions, paid events, business-profile claiming, an AI chat assistant, and admin tools. Caddy handles data enrichment, scraping, content generation, and location search. It cannot make a production change until a person approves it. I wrote every line of production code.

Challenge

TransUnion's credit report signup funnel was underperforming. Apex Growth hired me to build and test a new signup flow for TransUnion and NerdWallet. The flow uses knowledge-based authentication and handles PII. At roughly 6,000 transactions a month, it also had to meet PCI DSS v4.0 requirements.

What We Built

I took the designs through to a complete signup flow and handled the engineering work for PCI certification. That covered two penetration tests, monthly audits, remediation, and control documentation. Weekly A/B tests ran across desktop and mobile. One set of form-field changes lifted conversion 54% across all devices over 14 days, including 68.6% on mobile and 40% on desktop.

Challenge

Apex Growth manages roughly $6 million in media spend each month for about 20 client accounts across five ad platforms. Account teams were still stitching platform exports together by hand. They needed one reporting system they could trust.

What We Built

The BigQuery platform shipped in four months. I was the sole engineer. It brings Google, Meta, Microsoft, LinkedIn, and TikTok data into one place. Automated checks monitor quality and freshness. Account teams use internal AI tools to analyze campaigns and run weekly reviews.

Challenge

The firm had built a different solution for every client. Maintaining those one-off builds did not scale, and teams still spent hours pulling performance data by hand. They needed one internal platform.

What We Built

One data model gave data from Google Ads, Facebook Marketing, BigQuery, Snowflake, Google Sheets, and Slack the same structure. Non-technical marketers built their own dashboards with drag-and-drop controls, then used AI tools to ask ROI questions in plain English. I led the frontend and backend rebuild. That work led directly to the Apex Growth data platform.

Challenge

The idea had already found an audience on social media. The pitch was simple. Upload one selfie and get back a themed portrait pack. The product had to take payments, run GPU jobs, and deliver the finished images within minutes.

What We Built

I built the product from scratch, including user accounts, payments, the full AI generation pipeline, and real-time delivery. More than 80 themed packs each produced 20 unique images from one selfie. I learned the AI and ML stack on the job, then set up GPU infrastructure, batch processing, and edge functions. Customers could buy a pack, start a job, and download their portraits within minutes.

Challenge

The founders described SWURL as "LinkedIn for creatives," an invite-only network for senior creatives. The product had been paused for a year. They needed a technical lead to recover it and finish the web experience. A mobile developer was working in parallel and needed a documented API.

What We Built

I recovered the paused codebase and brought it to a full MVP with onboarding questionnaires, member profiles, a filterable directory with discovery and matching, messaging, and expert insights. That included designing the data architecture, setting up authentication, billing, separate staging and production environments, and building and documenting the API used by the mobile developer.

Challenge

ALYVR, a 3D architectural visualization startup, was close to launch. Its interface still made sense to engineers, not the architects and designers expected to use it.

What We Built

A UX designer and I redesigned the interface. We improved account management and performance, fixed longstanding bugs, and made the whole product responsive. ALYVR launched and started onboarding clients.