Hyper-Personalisation at Scale: Engineering the Future of the F1 Fan Experience
Upshift Ventures is a London-based startup dedicated to revolutionising fan engagement. By partnering with Pluralit AI Solutions, they sought to build GenF1, a hyper-personalised content platform designed to serve millions of users worldwide.
Together, we designed and deployed three end-to-end GenAI solutions, moving from PoC through Pilot and Pre-Production. Each solution was co-developed with the client’s editorial, product, and engineering teams using weekly sprints, user-driven feedback, and strict quality benchmarks.
The GenF1 initiative was a high-stakes technical PoC, designed to prove technical feasibility and experience differentiation to potential investors.
The objective was to deliver a fully functional Demonstrator: an adaptive app that responds to individual user preferences, within a strictly limited time and budget perimeter.
THE PROJECT
The primary goal was to build a Demonstrator (PoC) to validate whether an AI-native app could handle complex personalisation at scale.
Upshift Ventures required a solution that could overcome two hurdles to prove the apps’ commercial viability.
During this phase, we aimed to address two significant technical risks prior to full-scale production:
-
Architectural Scaling: we needed to prove the system could scale to millions of users with manageable operational costs, avoiding the slow speeds and high expenses of real-time, per-user web searches.
-
Temporal Accuracy: to simulate a true “live” experience, the AI had to be “blinded” to race results during simulations to prevent it from hallucinating information about future incidents.
THE CHALLENGE
THE SOLUTION
-
The Tri-State Knowledge Model: We validated a three-tier architecture that separates general history in Postgres/pgvector from high-frequency signals, such as lap leaderboards and pit stops, managed in separate Race-Day Simulation Tables.
-
Intelligent Content Routing: the system prioritises a proprietary Editorial Foundation (expert-written content establishing the brand voice) before querying the Curated Knowledge Base and, finally, external web sources.
-
The Optimisation Pivot: to ensure commercial scalability, we implemented “Cached Pooling”. By sharing search results across user segments, we drastically reduced API consumption and token costs.
-
Hyper-Personalisation Proven: we delivered distinct experiences for micro-segments, accurately catering to various personas such as a “McLaren technical fan” vs. a “Hamilton lifestyle fan”.
Pluralit AI Solutions designed and delivered a LangGraph-powered PoC that unified F1-specific data requirements with an advanced agentic framework:
RESULTS & BENEFITS
The delivery of the GenF1 Demonstrator transformed Upshift Venture’s strategic position:
-
Technical Validation: we proved that the architecture can support millions of users while keeping costs controlled through advanced caching and web consumption models.
-
Superior User Experience: expert testers confirmed the output was significantly more sophisticated and better-toned than generic AI, remaining accurate and grounded in the “GenF1 voice”.
-
Fundraising Success: the PoC served as a powerful tangible asset, moving investor conversations from abstract ideas to an engaging, proven experience with clear path to production.
-
Operation Roadmap: we established definitive plan to transition from demonstrator to a full Service App via automated ETL pipelines and licensed content ingestion.
The GenF1 Demonstrator was more than a technical exercise; it provided Upshift Ventures with a tangible asset that moved investor conversations from abstract ideas to a proven experience. By tackling cost-efficient pooling and data integrity from the start, we ensured that the transition to a full production service is a matter of execution, not just theory.
At Pluralit AI Solutions, we build for the long term. We combine deep engineering with the strategic foresight needed to turn a roadmap into a market-ready product.
Whether you are navigating a complex data environment or scaling for millions of users, we provide the architecture and the expertise to deliver.
Ready to build your next generation of AI?
