On-premise AI where consumer data can't leave
Confidential — Insurance Lead Marketplace
A rapidly scaling lead-generation operation spanned media buying, aggregated feeds, owned campaigns, and programmatic acquisition, with a team across three countries and a constellation of platforms for distribution, routing, validation, and fraud detection. AI could sharpen scoring, quality, and routing — but in a regulatory environment where consumer data leaving the building is an existential risk.
The work started inside the business: the economics of a lead, the margin at each handoff, the exposure at every data touchpoint. We mapped the end-to-end lead lifecycle and audited more than twelve integration points.
The defining decision was building entirely on-premise: custom hardware specified for local LLM inference, tuned to the workloads the business actually runs. On top of it, a roadmap of three model designs — predictive lead scoring trained on proprietary conversion data, automated quality assurance that catches issues before leads reach buyers, and intelligent call routing that optimizes match quality in real time. No cloud AI APIs. No data leaving the building.
- →On-premise AI infrastructure design and custom hardware specification
- →Local LLM deployment architecture
- →Predictive lead-scoring model design
- →Automated AI-driven quality-assurance pipeline
- →Intelligent call-routing optimization
- →Data-sovereignty and compliance framework, plus scaling roadmap
“What set Rodrigo apart was how quickly he understood our business. Within weeks he was speaking our language — regulatory nuances, market dynamics, all of it. He didn't just bring AI expertise; he became an expert in what we actually do.”