SCENARIO · TOP PROFILE
A rare profile deserves the right mission.
Top candidates make their decision in two weeks. Cobalt instantly identifies the accounts that book this kind of profile, prepares a targeted proposal for each one and gets you in front of the right decision-maker before the competition.
WHAT'S AT STAKE
Mis-positioning a rare profile costs their credibility for six months.
A nuclear CFD engineer turns down three generic approaches and stops answering. A senior ML who receives a run-of-the-mill "data scientist" role disengages. The precision of the first proposal decides whether this profile will work with you — or with your competitor.
2 wks
decision window for a rare profile
6 mo
credibility lost after a poor positioning
80%
of the decision is on the 1st proposal
3
well-targeted accounts are enough
THE PLAYBOOK
From CV received to interview at the right client.
A rare profile lands in your pipeline. Here's how Cobalt places them on the right mission in 5 days instead of 3 weeks.
Fine qualification
Cobalt extracts skills, project contexts, tech stack and constraints from the CV + interview. Way beyond keywords.
Semantic account search
Cobalt queries your history: who booked an equivalent profile in the last 18 months? Which accounts are currently hiring?
3 targeted proposals
One proposal per priority account, written in the client's domain vocabulary. Branded skills dossier, Business Manager signature, explicit subject line.
Interview confirmed
Meeting scheduled with the account that responds. Your consultant meets the right decision-maker — not an HR filter.
WHAT MAKES IT POSSIBLE
Precision, not volume.
Placing a rare profile isn't about sending 50 emails. It's about sending the 3 right ones, well-written, at the right time, to the right people.
Semantic search
Forget keywords. Cobalt understands "nuclear CFD ≈ high-constraint fluid mechanics" and surfaces compatible accounts.
Historical account mapping
Cobalt finds in 2 seconds who already booked this profile type — in your portfolio, over the last 3 years.
Domain-aware proposals
The AI picks up the exact vocabulary from past client needs. An aerospace CIO immediately sees that we speak their domain.
Credibility protection
Alerts if a proposal targets an already-solicited or ill-calibrated account. You preserve the candidate's — and the client's — trust.
THE IMPACT
Rare profiles placed — not lost.
5 d
avg time to placement
×2
interview conversion rate
–80%
mis-targeted proposals
90%
of rare profiles placed
Frequently asked questions
How does Cobalt recognize a "rare profile" without me telling it?
Cobalt compares every new profile against your base and the market. A profile whose skill combination is rare (top 5% of your history) is automatically flagged as "rare" and handled with a priority playbook.
What if I have no account history for this type of profile?
Cobalt supplements your history with market data: who posted equivalent roles in the last 12 months, which accounts mention these skills in their communications. We rarely start from zero.
Can I still position a profile on an account Cobalt hasn't recommended?
Of course. Cobalt's recommendations are suggestions — never barriers. You keep final say on the accounts where you want the profile presented.
Won't AI-generated proposals feel impersonal?
Quite the opposite — the AI leans on the actual phrasings from past needs at the target account. CIOs we interviewed tell us: "it feels written by someone who actually knows our domain." That's the intended effect.
Does it work for very senior / executive profiles?
Yes — same logic, even more value: executive profiles demand hyper-precision, which Cobalt delivers natively. See also our /blog/ia-recrutement-cadres-dirigeants-executive-search article.
How many rare candidates can I handle simultaneously?
No limit. The playbook runs in parallel for each profile. If you get 10 a week, Cobalt handles 10 a week — no quality drop.