AI and data hiring,
done properly.

JL Recruitment — a boutique search practice for AI, machine learning and data teams. The people you want aren't sitting on job boards or reading job ads. We know how to reach them, and how to identify if they can deliver.

[email protected]

Replies within one working day · searches taken on selectively

Start a search See the process

Scroll to run

search — Head of AI Closed

Series B applied-AI · Sydney 24 days

184

Mapped

Screened

Introduced

J. OkonkwoHired

A. ReyesFinal

M. HalliwellSecond

S. DevlinSecond

ResultOffer accepted

Head of AIMachine Learning EngineerResearch ScientistData EngineerChief Data OfficerMLOps EngineerApplied ScientistAnalytics EngineerFounding ML EngineerVP DataForward Deployed EngineerComputer Vision LeadData Product ManagerHead of ML Platform

Stage 01 · Brief

Half the brief is never written down

"Machine learning engineer" means five different jobs. Before any search starts we work out which one you actually need — research or production, platform or product, and whether the person is building the team or joining one.

intake — session 01

Q. Is this research, or shipping to production?

Q. What's the data actually like on day one?

Q. Who owns the model once it's live?

Q. What does failure look like at six months?

! Brief revised — this is a platform hire, not a scientist

✓ Signed off by hiring manager ▋

Stage 02 · Map

The people you want aren't applying

The best AI and data candidates are already employed. We map the Australian market and build relationships with passive, hard-to-reach talent — making hiring easier.

market map — 4 weeks out Live

184

In market

Reachable

Spoke to us

Product scale-ups72

Banking & insurance54

Research & spin-outs38

Global tech, Sydney20

ReadBudget is 12% under market — adjust or widen

Stage 03 · Shortlist

Four people, each with a written case

We provide tailored shortlists per requirement, with reasoning on every name: what they've actually shipped, what drives them, and what it will take to move them. Technical claims are checked before you ever see the CV.

shortlist — 4 of 31

Candidate 01 · J. Okonkwo Strong fit

+ Took two models from notebook to production at scale

+ Built the eval harness, not just the model

− Research background; hasn't managed before

− Will need a counter-offer plan; holds equity

Move: ownership and compute, not base salary ▋

Stage 04 · Close

In the room until the ink is dry

Offers fall over in the last mile. We handle the negotiation, the counter-offer, the resignation, and the notice period. We schedule Day 1, Week 1, and Month 1 check-ins, plus six-monthly check-ins, to ensure smooth integration into the team and prevent any surprises.

close — offer stage Accepted

✓ Offer issued day 21

! Counter-offer received day 23

→ Reframed on scope and ownership

✓ Accepted day 24

✓ Resigned, notice served clean

✓ Still in post at 36 months ▋

Coverage

The whole AI and data stack

Individual contributor → C-suite

01 · Research

Science

Research Scientist
Applied Scientist
Research Engineer
Chief Scientist
Post-doc into industry

02 · Build

ML & AI engineering

Machine Learning Engineer
AI Engineer
Software Engineer
MLOps / ML Platform
Inference & infrastructure
Forward Deployed Engineer

03 · Foundations

Data & analytics

Data Engineer
Analytics Engineer
Data Scientist
Data Product Manager
Data governance

04 · Lead

Leadership

Head of AI / ML
VP Data
Chief Data Officer
Chief AI Officer
Founding engineer

Market

What these people actually cost

Sydney · base salary · 2026

benchmark — indicative bands Updated Q3

Data Engineer · senior$150–190k

Software Engineer (AI)$160–210k

ML Engineer · senior$170–220k

Head of ML / AI$220–290k

VP Data / CDO$280–380k

NoteEquity and compute access move candidates as often as base

Indicative Sydney bands in AUD, permanent roles, excluding super, equity and bonus. Ask for the benchmark on your specific role and stage.

What we screen for

Four things a CV won't tell you

04 checks · every candidate

Shipped, not just trained

Plenty of people can train a model. Far fewer have kept one alive in production, with monitoring, retraining and someone paged when it drifts.

"What broke after it went live?"

Stage fit

A frontier lab and a Series A product team need different people. Big-company pedigree means little if the infrastructure they relied on won't exist here.

"What did you build yourself, and what was handed to you?"

What actually moves them

In this market it's rarely just cash. It's compute, data they can't get elsewhere, autonomy, publishing rights, and who they'd be learning from.

"What would you need to say no to us?"

How they work with everyone else

Most data hires fail at the seams — with engineering, product and the people who own the numbers. We test that interface, and take references seriously.

"Who had to trust your model, and did they?"

Track record

Measured by who's still there

Last 12 months · AI & data search

50⁺

placements

1.6^yrs

average tenure

days, on average, brief to placement

retained searches, delivered in partnership with trusted collaborators on executive search mandates

Kind words

What people say afterwards

Recommendations · LinkedIn

Placed candidate

"Positive, responsive and supportive the whole way through — he placed me in a data engineering role and stayed in touch long after I started."

Senior AI Engineer