Nexarank connects proposal criteria, weighting and regional constraints so a reviewer can explain a recommendation. I contributed research, the user flow, pitch development and delivery, and formative testing.
- Context
- Ontario data-centre proposal prioritization
- Decision model
- Six criteria · 100 weighted points
- Evidence
- Five formative proxy-user sessions
- Output
- Team prototype, user flow and pitch
What I contributed
- Researched the problem and decision context
- Developed the user flow
- Contributed to pitch development and delivered part of the pitch
- Participated in formative proxy-user testing
Methods & making
Four-person Group 4 team. Individual contributions are described separately from collective work.
A high rank still needs a reason.
The hackathon brief asked how a workforce planning board could compare proposed data centres across jobs, community value, energy capacity and data sovereignty. The challenge was to make those competing priorities understandable within a short review, while keeping the final judgment with a person.
Test whether people can explain the result.
The team’s rapid formative study used moderated think-aloud sessions with five proxy users. The support report records requests for visible score construction and difficulty comparing close totals.
- 5
- Proxy-user sessionsFormative research, not policy validation.
- 10–15 min
- Reported session lengthFocused on interpretation of the prototype.
I participated in testing alongside my research, flow and pitch work. The findings are reported at team level.
This small proxy sample did not establish policy accuracy, ranking accuracy or real-world decision performance.
Decision 01 · Expose what carries the weight.
The confirmed model gives jobs the greatest weight, balances workforce equity with energy fit, and keeps investment, training and data sovereignty visible as separate considerations.
The team’s formative report describes distrust of rankings whose construction was hidden. Participants wanted to inspect the criteria behind the result.
Show the six dimensions and their weighted contributions, then provide a plain-language explanation of how the score works.
The 25 / 20 / 20 / 15 / 10 / 10 model is confirmed. The prototype’s historical example totals and percentages are not a validated scoring engine.
Give the reviewer a route through the evidence.
The flow links a ranked overview to regional capacity, detailed scorecards and a dedicated comparison view. It also makes proposal review, missing information and a human decision part of the wider process.
I developed the user flow as part of the team concept.
This original flow retains its early approval labels and “<10 points” branch. Later screens use different recommendation labels; the artifacts do not establish a resolved production rule.
Move from the ranking to its context.
Explore the original dashboard and capacity view. Each adds a different reason to pause before relying on the overall rank.
Ranked overview
The dashboard gathers proposal status, category comparisons and regional context in one review surface.
Original team dashboard. Historical mock totals differ from the detail views, and some displayed percentages do not match their point ratios.Let a reviewer start with the queue and choose which proposal or dimension needs closer attention.
Original team interface. Company names, locations and figures are illustrative prototype content, not actual approvals or verified proposal data.
Decision 02 · Compare the differences directly.
The formative report describes people pausing on close scores when no direct comparison was available.
Give two proposals a shared comparison surface for inclusion, investment and energy fit, supporting an explanation of the trade-off rather than a rank alone.
The side-by-side view is an original design artifact. Its “wins” label and example totals do not demonstrate a correct or tested selection algorithm.
A clearer decision-support concept.
The team produced a high-fidelity prototype, a review flow and a five-minute pitch, with five proxy sessions informing the emphasis on explanation. My contribution connected research, the user’s route through the evidence and how we presented the concept.
What the original artifacts establish.
The final weights describe the intended decision model. The untouched prototype screens document the design work. Their illustrative totals, percentages and labels contain inconsistencies: for example, the support document’s category points sum to 61.5 while its example total reads 83. The portfolio therefore treats the screens as historical interface evidence, without presenting them as verified calculations or an operational government approval system.