FeaturesEVO JD
Define the role

EVO disagrees with your role before looking for anyone.

It interviews whoever asked for the position, points out what does not add up, and refuses to write the description while a blocker is still open. You leave the conversation with the role corrected, and with the argument ready to take to the hiring manager.

The problem

Why this exists

How it works today

The description arrives finished, or gets written in fifteen minutes by copying the last similar one. The search tool takes what it was given and goes looking. Four weeks later the pipeline is empty and nobody can say which requirement is to blame.

How it works with EVO

Before the first search, it challenges each weak point with evidence and a practical consequence. What was a matter of opinion ("I think they need English") becomes a conversation about data ("fluent English cuts 3% off the remaining funnel, is it worth it?").

How it works

EVO JD from the inside

1Briefing and interview

It asks until it understands the role, not until a form is full.

There are eight dimensions of discovery: business and context, the problem to solve, permanent scope, short-term mission, governance and autonomy, the current environment, technical and behavioral profile, market and feasibility.

Each answer you give becomes a captured fact, tagged with the dimension it belongs to. The panel shows how many dimensions already have substance and which are still empty.

And it writes nothing in this phase, on purpose. Polished text written too early becomes an anchor: you end up defending the sentence instead of the role.
EVO JD · phase 2, Interview
✓ 1 Briefing2 Interview3 Challenge4 Drafting5 Audit

Got it. So the pain is not a shortage of technical people, it is that nobody can see the status of the queue. Let me record that:

Problem: projects miss deadlines and there is no visibility of status: a governance failure, not a technical capacity one
Governance: the CEO is the Product Owner and holds prioritization; the role executes, it does not define what to do
Scope: leads 3 discovery analysts, currently with no lead
JD score18/100
Dimensions3/8
The 8 dimensions of discovery
Business and context2 facts
Problem to solve1 fact
Permanent scope3 facts
Governance and autonomynow
Short-term mission
Environment and as-is
Technical and behavioral profile
Market and feasibility
2Challenge

Here it disagrees, with evidence, a consequence and a way out.

Every challenge has the same anatomy: the claim in one sentence, the argument in three paragraphs (what we observed, what it means, what it will cost) and a concrete suggestion. You accept, propose something else, or take the risk and state why.

A blocker stops the drafting. A warning does not stop it, but it is recorded. And there is one challenge, discrimination and legal risk, that has no option to take the risk.

None of this is the model’s opinion. It is 14 deterministic detectors running in code; Claude only drafts the argument after the rule has fired.
EVO JD · phase 3 of 6, Challenge
✓ 1 Briefing✓ 2 Interview3 Challenge4 Drafting5 Audit
BLOCKERB2 · incoherent seniority
You asked for a "Senior Specialist", but described a management role.

Of the responsibilities we captured, six are people leadership: defining what each analyst works on, running development plans and 1:1s, and owning the team’s capacity.

The practical consequence: you will attract a strong IC who turns down the management part, or a manager who accepts and then asks for the title back. Both scenarios cost you a cycle.

Suggestion

Reclassify as Team Leader and keep the scope. The range goes up ~25%: salary intel estimates R$ 18–24k for a Discovery TL in São Paulo, against R$ 14–19k for a senior specialist.

"I want your confirmation on both blockers before I write. I am not drafting this with the wrong seniority."

3The argument for the manager

You do not go back to the manager empty-handed.

For each challenge, EVO assembles the text ready for you to send to whoever asked for the role, in the structure that works in a negotiation: acknowledgment, data, proposal, and the consequence of not changing.

And if the manager prefers to answer directly, they get a link. No account to create, nothing to install: they answer the agenda, and the answer comes back as a confirmed fact that reprocesses the description.

Challenges · B2 · argument ready
2Open
5Accepted
1Risk accepted
Argument for the requester

"Rafael, the scope you described makes complete sense, and that is exactly why it does not fit under a senior specialist title. Of the responsibilities we captured, six are people management: development plans, 1:1s, performance plans and team capacity.

If we open it as a specialist role, we will attract strong technical people who turn down the management part, or a manager who takes the title and asks for a correction in six months. The range for a Discovery Team Leader in São Paulo is R$ 18–24k, against R$ 14–19k for a senior specialist.

I propose we reclassify and keep the scope. If budget blocks it, the alternative is to take people management out of the scope, but then we need to decide who looks after the three analysts."

acknowledgment → data → proposal → consequence · source: seniority canon + salary base

Under the hood

The 14 detectors run in code, not in the prompt

Each one fires on a deterministic rule. The model comes in only afterwards, to draft the argument, which means the rule does not vanish when the model changes version or mood.

B2Seniority incoherent with the scope described
B4Inflated must-haves, the funnel closes
B6Salary range below the market estimate
B7The vision of success is not measurable
B8Governance and autonomy left undefined
B9Discrimination and legal risk, with no option to accept it
A1Jargon used without expansion on first occurrence
A2Cliché in the behavioral profile ("proactive, hands-on")
A7Client history suggests an undeclared negative signal
+5Impossible requirement, internal contradiction, tooling gap, scope with no owner, unrealistic deadline

Searchability and completeness are also calculated in code: arithmetic, not an estimate from the model.

What comes next

How it fits into the rest of the journey

Bring a role that is hard to fill

It takes a real role to see EVO work, not a canned example.

Start using it