01
Role Design
Define the outcome, the work, the Success Profile and what good looks like in the first 90 days, before opening a search.
Best when: The role is new, fuzzy, multi-region or inventing itself in the job post.
For companies
Build the team your next chapter requires.
PathwaysX works with high-growth technology companies and enterprises making consequential hires in AI-adjacent functions, where the role definition, the evaluation method and the shortlist all have to be built from scratch.
01The engagement
A search runs in four movements, from outcome to offer.
Every stage produces something you keep: a calibrated role definition, a weighted profile, an evidence rubric. Even if the search stops, the design work stays yours.
- 01
Calibrate
Week 0
Outcome → Capability → Talent Model → Role. We interview stakeholders, align on what success looks like in 12 months, and write the role from the outcome backwards.
- 02
Source & evaluate
Weeks 1–2
A human Talent Partner runs the market: direct sourcing, network activation and structured screens scored against the profile, not gut feel.
- 03
Present finalists
Weeks 2–3
Three to five finalists, each with a Decision Pack: profile scores, evidence sources, trade-offs and risk flags.
- 04
Support decision
Offer & beyond
We sit in on debriefs, pressure-test the decision, support offer design and stay close through the new hire's first 30 days.
02Offer stack
Four ways to work with us. Start with role design and add from there.
We separate the problem of defining the work from the decision to hire. That keeps a workshop from becoming generic training, and a search from opening against a fictional job description.
02
Critical Hire
Find and close the right person for a consequential role, with an evidence shortlist, decision support and close support.
Best when: Getting the hire wrong would cost more than the fee.
03
Workshops
Facilitated sessions tied to a real problem: shape roles, align leaders or prepare a team for an AI-enabled operating model.
Best when: The blocker is alignment on the work, not access to résumés.
04
Implementation Team
Assemble a time-boxed team to build and ship the outcome, not a permanent staff-augmentation bench.
Best when: The work needs a delivery team rather than one permanent hire.
How it usually flows
Role Design → Critical Hire when you are ready to hire → Workshop when the room needs alignment → Implementation Team when a crew needs to ship.
03The Success Profile
Every role gets a weighted Success Profile before we open the search.
The profile states what the role actually requires, how much each dimension matters, and what evidence convinces us. Interviewers score against it; the Decision Pack reports against it. Everyone grades the same paper.
| Dimension | Weight | Evidence that convinces us |
|---|---|---|
| Business & product judgement | 25% | Has owned a P&L or roadmap through a pivot; can narrate a call they got wrong and what changed after. |
| AI fluency in practice | 20% | Can walk through a system shipped with agents or models in production: constraints, evals and failure modes, not a tool demo. |
| Operating range | 20% | Has built process from zero in an ambiguous function; references describe how they operated without a map. |
| Cross-functional influence | 20% | Has established working agreements across product, engineering and the business function the role serves. |
| Team leadership | 15% | Has hired and grown a team; can name someone they developed and where that person is now. |
Weighted profile vs. one candidate
Threshold 7.0 · Illustrative
- Business & product judgement25%8.6
- AI fluency in practice20%9.2
- Operating range20%8.1
- Cross-functional influence20%5.9 · flag
- Team leadership15%7.8
Strong builder, below threshold on cross-functional influence, raised in the Decision Pack before the offer stage, not after.
First 90 days: what good looks like
- A written 90-day operating plan the candidate brought to the table, not one we wrote for them.
- Two working relationships already established across functions the role depends on.
- One consequential decision made and documented using the judgement the profile weighted most heavily.
- A measurable operating baseline the team agrees not to weaken.
04Evaluation & Decision Pack
Two lenses on every finalist, then a pack you can interrogate.
Role readiness: can they do the job as designed? Scored against the Success Profile with cited evidence.
Future potential: where do they compound? Range, learning velocity and how they operate without a map.
Evidence
Profile scores tied to interview, work-sample and reference evidence.
Trade-offs
What each finalist gives you, and what you give up by choosing them.
Open questions
What remains unproven and the next question that would settle it.
Risk flags
Material concerns, their source and how to test them before offer.
Decision Pack · Finalist 02 of 04Specimen
Candidate A-84
Principal Applied AI Engineer
Search day 16 · Scored against Success Profile v2
- AI fluency in production9.4
- System architecture9.1
- Operating range8.8
- Cross-functional influence7.6
Evidence sources
- · Production eval-suite walkthrough
- · Work sample: agent failure-mode review
- · 3 backchannel references
Identified trade-off
Deepest evals fluency in the cohort; least exposure to regulated enterprise rollouts.
Open question
How do they sequence governance review without slowing the shipping cadence?
A human Talent Partner presents every pack. No black-box shortlists, no algorithmic rankings you can't inspect.
05Why teams bring us in
The moment is usually the same: the role matters, the playbook doesn't exist.
Sometimes it's the first AI-adjacent leadership hire. Sometimes an internal promotion created a gap nobody has hired for before. The common thread is that a miss is expensive and the standard process would be flying blind.
Consequential roles
Hires that shape the operating model. A miss costs two quarters, not two weeks.
No existing playbook
The job description would be fiction; the interview loop would be vibes.
Thin, hybrid market
The right profile combines disciplines and won't surface from a keyword search.
Operating cadence
Every two weeks: search status, who we spoke to, what the market is paying and what we would change. Each quarter: a tailored industry report on how AI is changing the work, roles and compensation.
Start a conversation
Bring us a role that doesn't have a job description yet.
A 30-minute calibration call is usually enough to tell whether the search is one we should run, and how we'd run it.