Requesty
Case Study|Enterprise AI Gateway

How Soccerment Builds Its AI-Native Football Intelligence Platform with Requesty

Soccerment s.r.l. · Sports Technology · AI-Powered Football Analytics · 330+ clubs, leagues and federations served

Days
Not weeks to ship
100%
Cost visibility
Zero
Provider lock-in
1
Gateway for all AI dev

We are a lean team competing with the largest players in football data. Requesty lets us build with AI like a company ten times our size, while keeping the cost discipline our investors expect.

Aldo Comi
Aldo Comi
Co-Founder & CEO, Soccerment

About Soccerment

Soccerment is an AI-powered football intelligence company headquartered in Milan. Soccerment applies advanced analytics and machine learning to player performance, scouting and match analysis. In 2025 Soccerment integrated SICS, an Italian leader in professional video analysis with over 20 years of experience serving leagues, federations and clubs across Europe. The combined group offers the SICS+ product suite: Atlas for scouting and match analysis, VideoCode for event coding, and XSEED, AI-powered wearables for physical performance tracking.

Soccerment is building SICS Atlas as an AI-native platform from the ground up. Large language models sit at the core of the product, powering automated scouting insights, natural-language data exploration and intelligent match reports, while the engineering team relies on AI-assisted development every day. For a venture-backed company competing with the largest names in football data, the speed and economics of AI are not a nice-to-have. They are the strategy.

The Challenge

Soccerment made an early bet on AI-assisted development to build Atlas, putting coding agents and AI tools at the centre of the engineering workflow. But as usage grew across developers, tools and model providers, a lean engineering team and a finance-driven leadership ran into a new class of problems:

  • Model fragmentation. Different coding tasks call for different models, and juggling separate provider accounts, tools and billing relationships was slowing the team down.
  • Provider dependency. Tying the entire development workflow to a single LLM provider meant exposure to outages, pricing changes and capability gaps, with Atlas's delivery timeline directly at stake.
  • Opaque AI economics. AI development spend was scattered across providers and tools with no unified view, making it hard to measure the return on the AI-first engineering strategy and to report credible numbers to the board and investors.
  • Engineering overhead. Every new model or coding tool meant another account, another configuration and another bill. An unacceptable tax for a small team racing against far larger competitors in football data.

Why Requesty

One API, every model

A unified, OpenAI and Anthropic compatible endpoint lets coding agents and AI development tools route through the gateway transparently. The team can test, switch or combine models without touching their workflow.

Intelligent routing and failover

Each coding task is routed to the best model for the job, and traffic shifts automatically if a provider degrades. Developers are never blocked and the Atlas build never stalls.

Unified cost tracking

A single dashboard shows AI spend per developer, per tool and per model, giving a finance-led leadership team investor-grade visibility into the cost of building Atlas.

Fully managed, zero infrastructure

There are no gateway servers to deploy or maintain, so engineering time goes into football intelligence rather than plumbing.

In Their Words

Requesty became our control tower for AI. Every tool and every model we use to build Atlas runs through one gateway, and I can see the cost of our development effort in real time.

Aldo Comi
Aldo Comi
Co-Founder & CEO, Soccerment