AI Matching Engine
Under the Hood: How Trevo's AI Actually Works.
Most platforms call keyword search “AI.” Trevo’s matching engine genuinely scores every person across ten weighted dimensions, shows you exactly why each person ranked where they did, and adjusts based on real outcomes.
What It Looks At
Core and secondary skills matched against the shared taxonomy, with proficiency weighting
How deeply each required skill is held — not just whether it's present
Years of experience, sector exposure, and project complexity
Relevant industry and domain background for the brief
Working style, company-size fit, and team-context signals
Cost against budget, target margin, and market benchmarks
Geographic proximity, remote capability, and travel willingness
Current utilisation, start date, notice period, and engagement window
Required certifications held and current
Contribution to the client's B-BBEE and Employment Equity goals
And the weights aren't fixed. You can tune them per client, per engagement, per requirement. Your business, your rules.
What Makes This Different
Transparent scoring
You see the reasoning, not just the ranking.
Proactive suggestions
Matches surface the moment a new brief is created.
Adjustable weights
Because a banking client and a tech startup value different things.
A learning loop
Every engagement outcome feeds back to sharpen future accuracy.
The Two-Score Model
Two Numbers That Matter — and We Keep Them Separate.
Trevo produces two distinct scores on your Resources, and never conflates them.
The match score
88How well a Resource fits this specific brief — computed live across the ten weighted dimensions, in the context of the spec. Contextual, per-engagement, fully explainable.
Answers: “right for this brief?”
The Trevo Resource Score
Out of 5, with commentary
How a Resource actually performs and what they're like to work with — built from real interactions and engagement outcomes over time. Persistent, relationship-based, qualitative.
Answers: “right, full stop?”
Together they de-risk who you put in front of a client — something no keyword-search ATS can do.
Try It
Tune the Weights. Watch It Reorder.
Drag any weight and the ranking recomputes live — or try a preset: a banking client and a tech start-up weight the world differently. Expand a person to see their per-dimension contribution.
Total weight: 100% · scores are normalised, so any mix works.
Ranking 5 consultants for this client engagement — drag a weight and watch it reorder.
The Data Layer
Better Matches Start With Better Data.
AI CV parsing
Structured profiles extracted automatically — skills, certifications, and experience normalised on the way in.
Bulk import
Migrate a whole talent pool at once, with validation and de-duplication built in.
Guided manual entry
For the profiles you build by hand — guided capture with smart defaults.
All three routes map to one shared skills taxonomy — so search behaves consistently across your firm and your partner network.
See It Score Your Actual Talent Pool.
In a 30-minute demo we'll run the matching engine against your real roles and people.