Matt
Heiting.

one of AdGood's first two employees. i own and build the technology across the company, with AI integrated through it.

• lead AI integration end to end: Claude as the working interface, internal knowledge repos, and an agent that migrated the company off HubSpot and now builds and maintains Salesforce.
• build and run Snowflake ingestion, commission systems, insertion-order flow, and nonprofit profiling for inventory matching and lead generation.

my independent practice: AI integration, product engineering, and applied research. i cover product, infrastructure, and marketing.

• built and ran Cubby v1 on Kubernetes (GCP), with multi-tenant routing, per-user databases, and auth. now building persistent business apps that AI can use through declared tools.
• built multiple iOS apps, currently in private testing.
• took a Salesforce consulting agency end to end on Claude, including skills, hooks, tools, CLI, and Claude as the working interface.
• research AI tooling hands-on: agent systems, LLM integrations, and what actually works.

• owned the end-to-end rebuild of analytics ingestion from Kafka streams of raw ad server data, supporting internal and external reporting, sales, and lead generation.
• automated Looker report generation and brought AI-assisted SQL into the ClickHouse warehouse workflow.

• owned analytics across business, marketing, growth, and product.
• built predictive models for retention, email timing, new user growth, and subscriber growth.

built the BI stack from the ground up: a Redshift data warehouse, Looker, and the LookML and reporting the company ran on.

built Excel template products that automated recurring finance and marketing work. helped teams automate dashboards and reporting inside the tools they already used.

managed Apple's marketing budgets at the agency across product lines and campaigns.

rebuilt the finance close from a data standpoint. replaced slow, error-prone manual steps with Postgres databases and Python and SQL automations.