Case studiesCase 04
−80%
time spent preparing each athlete's plan, with closer follow-up and rigorous, measurable alerts
- Before
- Plans built by hand, rule by rule
- After
- A plan generated and explained, ready to approve
Performance Command Center
The performance command centre for elite athletes
Automated periodisation rules, readiness data and medical context read by AI, with the coach approving every decision in a single flow.
- Sport
- AI and integrations
- Team
- 3 people
Results
100%
Of plans and recommendations approved by the coach before they reach the athlete
1 flow
A single review flow for rules, AI and agents: accept, edit or reject
1 athlete
Per AI session, with isolated context so data never mixes between athletes
700+
Teams in the Iterpro ecosystem, which remains the athlete's channel
NOMAS prepares training plans for professional footballers, with athletes such as Diogo Dalot and Luis Díaz. That work relied on manual tasks and on knowledge concentrated in the coaches, a model that works but doesn't scale. Agiliz designed and is building the Performance Command Center: a platform that automates plan generation, turns each athlete's data into recommendations and always keeps the coach's judgement at the centre of the decision.
- Client
- NOMAS Perform
- Sector
- Elite sport, footballer performance
- Technologies
- Next.js, TypeScript, PostgreSQL, AWS in a European region, Amazon Bedrock AgentCore, Iterpro API, WHOOP API, WhatsApp Business Cloud API
- Duration
- Since July 2026, with phased deliveries
- Team
- Guilherme Ferreira, Lourenço Barros, Rui Pereira
01The opportunity
Plans built by hand, athlete by athlete
Preparing each athlete's plan took up high-value time from the coaches, time taken away from work on the pitch.
Rules applied one by one
Days between matches, minutes played in the previous match, return-to-play management: periodisation rules were applied manually, in a slow process prone to inconsistency.
Rich data that never became decisions
Wearables, readiness and medical context were there, but they weren't turned into training decisions in a controlled way. And there was no trail of which decision was made and why.
02The solution
- 01
The coach's rules turned into a system.
Training cycles by days between matches and minutes played, standard rules and exceptions per athlete. The platform generates the right plan and always shows which rule was applied and why.
- 02
Two-way connection to Iterpro.
Athletes, exercises, calendar and minutes played come in automatically, and the approved plan goes back there. The athlete keeps using the tool they already know.
- 03
Direct connection to WHOOP.
Recovery, sleep, strain and heart rate variability come in every day. A bad night's sleep now counts in the plan before the athlete arrives at training.
- 04
AI that analyses rules and metrics.
It cross-references WHOOP with readiness, load, questionnaires and medical context, and gives explained recommendations, with the data and criteria behind them.
- 05
AI agents for the coaching staff.
A coordinating agent hands out tasks to sub-agents: monitoring athletes, preparing daily briefings, supporting content. Each agent only accesses the data it needs.
- 06
The coach always in charge.
Plans, recommendations and agent proposals go into a single review flow. Nothing reaches the athlete without human approval.
- 07
Health data protected by design.
AWS in a European region, an isolated AI session per athlete, a data protection impact assessment and a full audit trail from day one.
03How we implemented it
Through MEDIR, as in every project. It's the same screen the client follows in the portal, phase by phase.
M Map
Completed
E Estimate
Completed
D Decide
Completed
I Implement
Phased deliveries · since Jul 2026
R Review
In progress
Map
How coaches built each plan: rules applied one by one and wearable data with no link to decisions.
Rules and data
Estimate
High-value coaching time spent building plans, athlete by athlete, with no trail of each decision.
Coaches' time
Decide
Configurable rules, AI recommending and the coach approving in a single flow, with health data protected.
Plan approved
Implement
Performance Command Center in phased deliveries, connected to Iterpro and WHOOP.
Since Jul 2026
- Periodisation rules per cycle and per athletesince Jul 2026
- Two-way connection to Iterprosince Jul 2026
- Daily WHOOP data in each athlete's plansince Jul 2026
- Explained AI recommendationssince Jul 2026
- Single approval flow for the coachsince Jul 2026
ReviewCurrent phase
Each delivery reaches production with value of its own, ready for new data sources and new models.
Ongoing
04The impact
The coach no longer builds each plan by hand. They open the platform and find, for each athlete, the plan generated by the rules, the AI's recommendations with their reasoning and a single button to approve.
The NOMAS methodology is now applied consistently to every athlete, with a trail for each decision. The knowledge that lived in the coaches' heads is now in the system.
NOMAS gains the capacity to look after more athletes without growing the coaching staff at the same pace. The architecture is ready for new data sources and new AI models.
Let's build something real.
No more pilot projects that never reach production. We start with the Operations Map: two weeks, fixed price, to find out where hours, money and clients are being lost. Then we build, integrate and stay alongside the team.