PROJECT STUDY
PrototypeFootball Coach AI
One place for training, matches, and team context.
An academy workspace connecting training plans, curriculum, matches, and team insights, with a source-grounded assistant and coach approval.
Explore the build
At a glance
- Stage
- Prototype
- Industry
- Sports
- Context
- Internal product
- Capabilities
- 6
- Technologies
- 9
- Year
- 2026
The challenge
A football academy's decisions depend on more than a list of drills.
Coaches need to connect the team's age group, recent training, available equipment, upcoming matches, and the knowledge behind a plan. When those records sit in separate tools, the context is difficult to carry into the next session.
Football Coach AI explores a connected workspace for that work. Its assistant uses team records and coaching documents to help a coach prepare and analyze, while the coach retains responsibility for approval and scheduling.
What we built
The implemented prototype brings administration, planning, and review into one web workspace:
Academy and team context
Organizations, teams, players, memberships, equipment, and facilities provide the records behind each plan
Training and curriculum planning
Coaches can review training drafts, approve sessions, and plan multi-week curricula. The planner supports schedule previews and rechecks constraints when the coach confirms
A grounded coaching assistant
Typed, permission-checked read tools connect the assistant to stored team, training, curriculum, match, and load information. Knowledge retrieval adds source references that the coach can review
Matches and observations
Fixtures, results, player statistics, and pitch-zone events give coaches a place to record and discuss what happened
Team and player insights
Analytics bring training balance, recorded load, and performance trends into a shared view. AI reviews use the information that has been recorded and identify gaps
Calendar and knowledge workflows
An internal calendar supports conflict checks. Document ingestion, indexing, and search support the coaching library; Google Calendar connection and export paths are implemented
Academy and team context
Organizations, teams, players, memberships, equipment, and facilities provide the records behind each plan
Training and curriculum planning
Coaches can review training drafts, approve sessions, and plan multi-week curricula. The planner supports schedule previews and rechecks constraints when the coach confirms
A grounded coaching assistant
Typed, permission-checked read tools connect the assistant to stored team, training, curriculum, match, and load information. Knowledge retrieval adds source references that the coach can review
Matches and observations
Fixtures, results, player statistics, and pitch-zone events give coaches a place to record and discuss what happened
Team and player insights
Analytics bring training balance, recorded load, and performance trends into a shared view. AI reviews use the information that has been recorded and identify gaps
Calendar and knowledge workflows
An internal calendar supports conflict checks. Document ingestion, indexing, and search support the coaching library; Google Calendar connection and export paths are implemented
The assistant drafts and analyzes. It does not independently schedule sessions, modify team records, or delete data. A coach with the appropriate role reviews and approves the next action.
How it works
Next.js, Go and PostgreSQL
Next.js and React provide the interface. A Go API owns business rules, authorization, and database access. PostgreSQL holds the operational records, while ChromaDB supplies a derived search index for coaching knowledge.
Permission-checked assistant access
The assistant combines permission-checked record access with document retrieval. Training drafts are validated before entering the normal review workflow. Background jobs support ingestion and calendar work, keeping those tasks separate from the interactive conversation.
Multiple organizations
The same platform supports multiple organizations, so access checks and team assignments are part of the implementation rather than something left to the interface.
Current stage
Prototype
- Research
- Prototype
- Pilot
- Production
This is an internal prototype with implemented frontend pages and corresponding backend modules. The repository includes Go tests, integration tests, and Playwright journeys. Project documentation records earlier local verification; those checks were not rerun as part of this portfolio update.
Read the full status
Live AI responses, Google sign-in, and Google Calendar acceptance depend on configured credentials and require separate validation. No production deployment, real academy client, athlete outcome, or commercial result is claimed here. Google Classroom, writing player assessments and development goals, and per-player attendance writes are outside the documented release.
Where this fits
The same pattern works when a team needs answers grounded in both operational records and internal knowledge. We connect AI assistants, business systems, and data dashboards around access rules and human decisions. Talk to our team about a workspace for your own organization.
KEEP EXPLORING
Another problem. A different approach.

Logistics Intelligent
A dispatch and supply-planning prototype for delivery routes, fleet disruptions, and retail replenishment across multiple warehouses.

Venue & Decor Visualizer
An event-design prototype with privacy-aware venue capture, labelled preview workflows, and proposal tooling behind a bilingual foundation.
Let’s connect the dots
Let's put these ideas to work.
Share a process, product, or problem you want to improve. We will work through the possibilities together.