Part 1 of 2
AI-powered athlete discovery
Scouts describe the opportunity first; AI search then surfaces the athletes who fit the brief.
Before: manual discovery - Open scout dashboard
- Browse athlete list
- Open athlete profile
- Match for the opportunity?
No — review the next athlete (back to step 2)
- Begin detailed evaluation
Fig. 1 Too many athletes to scroll
Scouts couldn't review hundreds of thousands of profiles for one opportunity, so real evaluation started far too late.
Fig. 2 Start from the brief
Scouts describe the opportunity first, so the AI search has real criteria to work from instead of guessing.
Fig. 3 Must-haves vs nice-to-haves
Hard requirements decide who's eligible; soft preferences only change the ranking. Flexibility is explicit, not implied.
Fig. 4 Search that leads into the work
Results open straight into the scout's existing evaluation. The AI finds candidates; people still decide.
After: AI-assisted search - Enter opportunity brief
- Set hard and soft criteria
- Review requirements
- Run AI search
- Review relevant athletes
- Open profile for evaluation
Fig. 5 Six clear steps
Brief, criteria, review, search, shortlist, evaluate, instead of an endless scroll.
Fig. 6 Relevance over speed
Reaching an unsuitable athlete quickly isn't success, so testing covers no matches, missing data and refining a search.
Read the full write-up
Problem
Scouts could not reasonably scroll through hundreds of thousands of athletes to find the right fit for an opportunity. Manually reviewing profiles placed too much effort before meaningful evaluation could begin.
The design challenge was to capture what the scout needed, distinguish essential criteria from preferences, and translate those requirements into a search the backend could support.
Hypothesis
If scouts define the requirements of an opportunity before searching, our integrated AI evaluation system can use those criteria to search across Talstrike and surface athletes who best fit the brief.
Separating essential requirements from flexible preferences should help scouts narrow the field without unnecessarily excluding promising talent.
Goals
- Reduce the effort scouts spend finding suitable talent so they can begin real evaluation sooner.
- Give athletes greater visibility when their profiles match an opportunity.
- Reach relevant athlete profiles in a few clicks after entering the opportunity requirements.
Measured by: Time to the first relevant profile, interactions required to reach it, and the share of searches that lead to detailed athlete evaluation.
What I did
- I worked with the backend team to understand the infrastructure and translate the steps needed for AI-assisted search into the interface.
- I investigated how scouts frame an opportunity, so the search would begin with their requirements and feel natural to their decision-making.
- I introduced a distinction between hard requirements and soft preferences, so scouts keep essential criteria while staying flexible where it makes sense.
My role
My responsibility was turning scouting intent into an understandable search experience: gathering requirements, expressing priority, and connecting discovery to athlete evaluation.
Backend collaboration set the technical context. The AI system supports candidate discovery; scouts keep responsibility for judging suitability.
Solution
Search is centred on the opportunity. Scouts provide their requirements before the AI search begins and mark which criteria are mandatory and which are preferences. The system then searches Talstrike for athletes who fit, so scouts can focus on relevant profiles.
The central change is moving requirement definition ahead of discovery, so scouts no longer repeat the same assessment across a huge athlete list.
Key decisions
- Gather requirements before searching, so the scout's opportunity gives the system a meaningful starting point.
- Separate hard requirements from soft preferences, so flexibility is explicit.
- Keep detailed athlete evaluation one step from the results, so search leads straight into the scout's existing work.
- Matching rule: hard requirements decide eligibility; soft preferences influence ranking.
Trade-offs
- Manual browsing gives direct control but makes scouts repeat their assessment for every athlete.
- Treating every requirement as mandatory produces a narrow search; treating every requirement as optional weakens relevance. The hard/soft split resolves that tension.
Validation plan
- Representative scouts complete the same opportunity brief through manual discovery and through AI-assisted search. Observe whether they can express their criteria, understand hard vs soft requirements, recognise relevant results and begin evaluation.
- Cover edge cases: no matches, incomplete athlete information and searches that need refining. Record relevance judgements alongside time and clicks — reaching an unsuitable profile quickly is not success.
Outcome and next steps
An opportunity-led route into athlete discovery, with requirements collected before search and flexibility expressed through hard and soft criteria.
The intended outcome is less effort before evaluating suitable athletes and more visibility for athletes who match. Feature-level gains will be reported against a baseline using the measures above.
Bringing the scout's decision criteria into the search itself made “what is a good fit?” the central interaction question.
Next, I'd study where scouts change a requirement after seeing results, and whether the hard/soft distinction helps them do it confidently.
