Footalara

Inspiration

Football fans can find endless news and raw statistics, but daily football games are usually isolated, shallow, or backed by inconsistent data. Footalara turns football history into a coherent daily habit through seven different browser games that share one validated football-content platform.

What it does

Footalara is a free browser product containing:

• Club Connections — connect two teams through a player who made official senior appearances for both.
• Road to the Trophy — reconstruct real historical standings and knockout paths.
• Daily Football Quiz — answer ten daily questions ordered from accessible to difficult.
• Guess the Lineup — recover the starting elevens from an iconic match.
• Guess the Player — identify a footballer from progressively revealed clues.
• Guess the Player II — find a hidden player through similarity rankings.
• Higher or Lower — compare famous players across historical and seasonal statistics.

The games reuse stable player and team identities, verified careers, historical fixtures, competition editions, search systems, and daily-session infrastructure instead of behaving like seven unrelated prototypes.

What was built during OpenAI Build Week

Footalara existed before the event. This entry is based only on the meaningful work completed between July 13 and July 21, 2026.

During Build Week I:

• built and integrated the complete Road to the Trophy daily game;
• built and integrated the complete Club Connections daily game;
• substantially expanded Higher or Lower player coverage and comparable metrics;
• reclassified and stabilized the 1,830-question Daily Football Quiz bank;
• migrated the shared football-content layer to a Sportmonks-backed canonical model;
• expanded historical databases and runtime player coverage;
• unified player search, popularity ranking, careers, lineups, and tactical positions;
• added deterministic snapshots, validation gates, atomic activation, rollback protection, and strict API-budget controls;
• improved search latency and player-career timeline quality.

The repository preserves the real dated commit history and clearly distinguishes earlier work from Build Week additions.

How I used Codex and GPT-5.6

Codex with GPT-5.6 was the primary engineering environment used throughout the Build Week extension. It was used to inspect a large existing repository, trace data and API contracts, plan integrations, implement React and FastAPI functionality, design canonical football entities, migrate provider-era paths, generate validation logic, run and interpret tests, and diagnose PostgreSQL, Redis, search, lineup, tactical-position, and career-history issues.

Codex was used across the engineering loop rather than only for isolated code generation: repository analysis, implementation, data engineering, migration work, validation, debugging, iteration, and documentation.

GPT-5.6 powered those Codex sessions. Footalara does not falsely present GPT-5.6 as a runtime dependency where it is not one; its role in this entry is the engineering and reasoning capability used to build and extend the product.

How it was built

• React 18 and Vite frontend
• FastAPI and Python backend
• PostgreSQL production database
• Redis cache and session support
• Docker Compose deployment
• Caddy production edge
• deterministic offline football-content package
• compact SQLite search artefact
• background worker
• Vitest, Testing Library, and Pytest
• Sportmonks administrative ingestion

Data quality and reliability

• Gameplay does not call the football provider at runtime.
• Public IDs remain stable and provider IDs stay in internal crosswalks.
• Missing statistics remain missing; the system does not invent zero values.
• Publishable lineups require both teams and exactly 22 unique starters.
• Club Connections requires verified official senior appearances.
• Generated revisions are deterministic, hashed, audited, and activated atomically.
• Failed validation cannot replace the previous working production revision.
• Provider ingestion uses a hard request ledger and immutable safety ceiling.

Challenges

The hardest part was not producing another isolated football game. It was extending an existing product while keeping seven game contracts, daily sessions, historical data, search behaviour, PostgreSQL state, Redis caches, and rollback safety coherent. Several seemingly simple data problems—such as loans, returns to former clubs, tactical wide positions, incomplete historical lineups, and semantically different statistics—required explicit validation rather than optimistic assumptions.

Accomplishments

The Build Week work produced two complete new daily games, a much broader shared player catalogue, stronger historical coverage, faster search, more accurate career timelines, and a safer content-publication system. The final product is a complete, runnable consumer experience rather than a technical proof of concept.

What I learned

The project reinforced that AI-assisted engineering works best when the agent is given hard invariants, real test feedback, and permission to challenge weak assumptions. Codex accelerated repository navigation and implementation, but the strongest results came from repeatedly validating generated work against data contracts, historical evidence, budgets, and product behaviour.

What's next

The next steps are broader historical coverage, additional daily formats, stronger social sharing, improved onboarding, and better player-progress analytics while preserving the same validation and offline-runtime principles.

Try it

Open https://footalara.com. For the fastest evaluation, start with Club Connections and Road to the Trophy, the two major Build Week additions. No paid account is required.

The repository README contains local setup, automated test commands, architecture, dated Build Week evidence, and the judge testing path.

Built With

• caddy
• codex
• docker
• fastapi
• gpt-5.6
• postgresql
• pytest
• python
• react
• redis
• sportmonks
• vite
• vitest