Lab component 11

A datasheet instead of an about page

Every professional bio is written to be maximally acceptable, which makes them all identical and none of them useful. This one names the constraints and the failure modes, so a bad fit can disqualify itself in thirty seconds instead of three weeks.

Self reported Written to be disagreed with

Operator specification

Jake Labate

rev 2026.08
Boston, MA
independent
Primary function
Search infrastructure
Operating mode
Build, then explain
Team size fit
1 to 12
Ramp time
Days, not weeks
01

Working characteristics

02

Operating parameters

Preferred input
An ambiguous, repeated, annoying problem that nobody has bothered to formalise yet.
Typical output
A typed model, a reusable framework, and something deployable. Rarely a slide deck on its own.
Failure signal
Will build the general system when the specific fix was what was needed. Ask directly if you want the narrow version.
Communication default
Direct and technically dense. Acts rather than asking clarifying questions where the intent is reasonably clear.
Feedback preference
Blunt and early. A late polite note costs more than an immediate uncomfortable one.
Needs from a manager
Context on why, and the authority to change the underlying system rather than only its symptoms.
Does not need
Daily check ins, or a task queue that has already decided the approach.
03

Fit assessment

Strong fit

  • Undefined problem space. Nobody agrees what the thing even is yet, and the first real deliverable is a shared model.
  • Repeated manual work. Something a person does forty times a month that could be a system.
  • Technical and non technical audiences. Work that has to survive both an engineering review and an executive summary.
  • Small teams with real leverage. Where shipping is measured in days and one person owning a whole surface is normal.

Weak fit

  • Fixed scope execution. Where the approach is settled and the value is in following it exactly.
  • Deep hierarchy. Where changing a system requires four approvals and the change is not the point anyway.
  • Pure relationship roles. Where the substance is account management and the technical layer is somebody else's.
  • Volume content production. Where the need is throughput on a defined template rather than a better template.

Self reported and deliberately unflattering in places. The point of publishing it is not to look good, it is to let the wrong engagement disqualify itself before either of us spends three weeks finding out.

Personality instruments are descriptive, not predictive, and a self assessment is the weakest form of one. Read the section above as a stated working preference, which is falsifiable by anyone who has actually worked with me, rather than as a measurement.

Hiring is a matching problem, not a ranking problem

Most professional self presentation optimises for maximum acceptability, which produces candidates who are indistinguishable on paper and mismatched in practice. Stating the constraints is more useful to both sides.

01

Filters cut both ways

A spec sheet that names the weak fits loses some conversations. Every one it loses was going to end at week six anyway.

02

Failure modes are more informative than strengths

Everyone claims the same strengths. What somebody reliably gets wrong is the part that actually distinguishes them.

03

Legible is not the same as agreeable

The goal is being easy to evaluate accurately. That is a different target from being easy to like.