Searching for talent

Search on Ewance is built around one assumption: you find candidates by what they shipped, not by what they said about themselves.

The filter set reflects that. You'll find no "years of experience" filter — that's a CV-shaped question, not a portfolio-shaped one. What you do find is filters that map cleanly to "what kind of work has this person actually produced."

The filter set

Discipline

The primary axis. Pick from:

• Engineering — mechanical, electrical, structural, aerospace, biomedical, chemical, civil.
• Software & data — back-end, front-end, mobile, ML/AI, data engineering, security, DevOps, distributed systems.
• Business & strategy — corporate strategy, market entry, M&A, ops, finance, consulting-shape work.
• Life sciences & medicine — clinical reasoning, public health, biotech commercialisation, lab design.
• Design — product design, UX, brand, campaign creative.
• Social sciences & policy — research design, programme evaluation, policy briefs.

You can stack disciplines for cross-functional roles ("software + product design" for a generalist PM-engineer hybrid).

Deliverable type

Independent of discipline. Filters by what kind of artefact the candidate has shipped:

• Written analysis — reports, market analyses, research papers, policy briefs.
• Slide deck — strategy decks, pitches, campaign concepts.
• Code repository — software projects with READMEs, tests, deployment notes.
• Design artefact — Figma files, wireframes, brand systems, CAD models.
• Mixed / multi-format — combinations.

If you're hiring a strategy associate, you want "slide deck" + "written analysis." If you're hiring a backend engineer, you want "code repository." Filter accordingly.

University and region

For in-person roles, visa scoping, or specific alumni preferences. The pool spans universities globally — Western Europe and North America are densest, with growing coverage in Latin America, Africa, and Asia.

Team format

What kind of collaborator are you hiring for?

• Solo-only — surfaces candidates whose portfolio is entirely individual work. Signal: "this person can carry a project alone."
• Team-experienced — surfaces candidates who've shipped at least one team submission. Signal: "this person can integrate with others."
• Mixed-discipline experience — narrowest filter. Surfaces candidates who've worked in cross-disciplinary squads. Signal: "this person can translate across professional vocabularies." Strong fit for larger orgs.

Challenge sponsor

If your team or an adjacent industry has sponsored a challenge, candidates who completed your sponsored brief surface here. This is the highest-signal filter for sponsoring recruiters — the candidate's first project on the platform is structurally close to the work you'd ask them to do.

Rubric outcome

The most subtle filter. Options:

• Accepted on first submission — signal: "got it right the first time." Bias: rewards candidates with prior exposure to the work shape.
• Accepted after revision — signal: "responded well to feedback." Often a better hiring signal than first-submission acceptance — it tells you the candidate can take criticism, integrate it, and ship.
• Both — default; widest pool.

Most experienced recruiters end up filtering for "both" or even leaning toward "accepted after revision" — first-pass perfection is a flat signal; responsiveness to feedback is a growth signal.

Availability

Three states: actively looking, open to conversations, not currently looking. Set this loose at the discovery stage and tighten it before outreach.

Reading a candidate's profile quickly

Once you have a shortlist, you'll review profiles. A good triage flow:

1. Headline read — discipline, top 1-2 challenges, university, year. 15 seconds.
2. Open one submission in full — pick the challenge closest to your role. Read the deliverable like you'd read a junior employee's first work. 5-15 minutes.
3. Skim the others — what's the range? Are they good at one thing or several?
4. Check rubric outcomes — first-submission vs. revision patterns tell you something about the candidate's growth profile.
5. Verify one credential — open a verification link in a new tab. Confirms the record is real. 30 seconds.

After steps 1-5 you'll have a strong sense of whether to message. Don't message before you've at least read one submission — generic outreach gets ignored, and you'll waste your own time.

Signals worth weighting up

These tend to be under-weighted by recruiters new to skills-based hiring:

• Range across deliverable types. A candidate who's shipped both a written analysis and a code repo demonstrates a flexibility CVs don't capture.
• Submissions outside their stated discipline. A business student who shipped a software prototype, or an engineer who shipped a strategy deck. Strong signal of intellectual range.
• Long-form revision history. A candidate who took three revision rounds to ship something genuinely hard often beats a candidate who flew through three easy challenges.
• Mixed-discipline squad experience. Pays back disproportionately for larger orgs.

Signals worth weighting down

• Pure quantity. Twelve easy challenges < three substantial ones. Read the work, not the count.
• University prestige in isolation. You can see the university, but the portfolio is the primary signal. Otherwise you'd be back to CV screening with extra steps.

Saved searches and alerts

On the free recruiter tier, search is unlimited but un-saved — every session starts fresh. On the paid tier (when it launches), saved searches and new-candidate alerts let you build a passive pipeline.

For now, the workflow is: open Ewance, run the search, message the candidates, close the tab.

Next

Once you have a candidate you'd like to talk to: From shortlist to hire.