They knew what they needed. The job was to go and get it.

Time to fill is measured from the brief to the person's first day of work. Outcome figures are the client's own, attributed to them by name, never our measurement. People are named as first initial and surname by house style.

The situation

Most of the companies on this page arrived with a goal and no job description. AIM arrived with the opposite. They had already worked out where the business was constrained, and it was in two places at once: not enough people selling, and not enough people building. They wanted three sales development representatives, one AI engineer and one machine learning engineer, and they wanted them quickly.

We include this case because the other three could give the impression that every engagement begins with a diagnosis. It does not. If you already know what you need, a discovery process is overhead, and the useful thing we can do is fill the seats fast and well. That is what happened here.

The team we built

• 3 sales development representativesFilled in 4 days

  

  Outbound capacity for a company selling AI services. Briefed, shortlisted, interviewed and started inside four days from the brief to the first day of work.

• 1 AI engineerFilled in 6 days

  

  A senior technical seat. This is the market where most agencies' timelines run to months, because the good candidates are employed and not looking. Six days.

• 1 machine learning engineerFilled in 6 days

  

  Placed alongside the AI engineer, on the same six-day clock. Engineering seats with us are scoped and priced separately from back-office roles, because they are a different hiring pool and a different price.

What a defined brief changes

When a client arrives with the roles already decided, the work is different, and it is worth spelling out what we did and did not do.

• We did not run a diagnosis. AIM had already located its constraint. Re-doing that work would have been a week of overhead for a conclusion they had reached. If the brief had looked wrong to us we would have said so on the call; it did not.
• We did split the brief into two searches. Sales development and machine-learning engineering are different talent pools with different screening. The SDR search was about outbound track record and hours coverage. The engineering search was about shipped work and the specific stack, and the interviews were technical. Treating them as one "five people" order would have slowed the fast one down to the speed of the hard one.
• We did keep the client's involvement to the decisions. Three candidates per seat, interviews scheduled, the client chose. Contracting, payroll and the start date were ours to handle. For a company hiring engineers while it is also hiring salespeople, the time not spent on logistics was the point.

The alternative most companies in AIM's position weigh is a technical recruiter on a placement fee, typically 20 to 25 percent of first-year salary, with a time to offer measured in weeks and a notice period after that. It is a reasonable route when you want the engineer on your own payroll from day one. It is not fast, and it does not come with a replacement guarantee that runs past the first few months.

Why the two numbers are different

Four days for sales and six for engineering. We publish both rather than averaging them, because the split is itself the credible part. A real operation produces different numbers for different role types. A flat "four days for everything" would read like marketing, and it would also be wrong.

Six days for an ML engineer is the more impressive of the two, and most readers would not know it. Senior technical hiring is where every agency's time to fill blows out, because the people worth hiring already have jobs and a notice period. Ours were contractors we had already headhunted and screened, with availability confirmed the week AIM briefed us, so the six days was matching and interviewing, not sourcing from zero.

Both numbers are measured from the brief to the person's first day of work. That is the only clock we use on this site.

What happened next

AIM reports that its output and sales numbers are up by at least 40% since the team started. That is AIM's figure, given to us by the client, and we present it as theirs rather than as something we measured. It is the first client-side result in this set of case studies; the rest of what we publish, seats filled and days to fill, is our own output.

What this case shows

• Execution without the consulting. When the brief is already right, we do not slow it down with a discovery process. Five seats, two role families, started within a week.
• Technical depth. AI and ML engineers are a different market from executive assistants and customer service. A staffing partner that fills both is not competing in the assistant lane, and it is the hardest of these four stories for a competitor to match.
• Different clocks for different roles. Four days and six days, published as two numbers because that is what actually happened.
• The client's number, attributed. We do not claim outcomes as our own measurement. AIM's figure is presented as AIM's.

Questions about this case