Four things said in the boardroom.
A few things that have come up more than once in conversations this month.
-
1
“We need AI engineers.” Quite a few companies still can’t explain what they mean.
We’re seeing a lot of businesses ask for AI Engineers when what they actually need ranges drastically.
And it keeps coming up.
“We need an AI Engineer.”
Great. What do you actually need them to do?
Sometimes they need a very good software engineer who knows how to use AI properly. Sometimes they need someone building agents, RAG and AI systems in production.
Those aren't the same person.
And sticking AI Engineer on the job title doesn't fix a muddled brief.
-
2
Contractors are being offered permanent jobs. Some are just leaving instead.
I've heard a few versions of this recently.
Contract roles get cut. A smaller number of permanent jobs are offered back.
Makes sense on a spreadsheet.
The kicker is that quite a few contractors became contractors because they wanted to contract.
Take away the flexibility and they don't automatically become permanent employees.
Some just walk.
-
3
AI budgets are becoming an engineering problem.
We’ve heard a lot of examples of engineers burning through enterprise AI credits in the first few weeks of the month, then waiting for the next allocation.
We've spent the last couple of years asking whether engineering teams should use AI.
I reckon the more pressing valuable question now is:
How do you actually run the thing properly without wasting money?
-
4
“What language do you code in?” is becoming a less useful hiring question.
This one keeps coming up in hiring conversations. Can they read unfamiliar code? Can they review something AI wrote? Can they work across the stack? Do they understand the actual business problem? Can they tell when the AI has produced bollocks? I'd probably rather know all of that.
Offshoring might not be as cheap as the spreadsheet says.
India is still a huge talent market. The interesting bit is what happens when everyone wants the same top slice of it.
Loads of businesses are looking at offshore engineering hubs again.
India is usually somewhere in the conversation. And there are obviously very good engineers there. But as always, everyone wants the good ones.
So you're not competing for “Indian engineering talent”. You're competing with every other company trying to hire the top end of the same market.
That pushes salaries up. Suddenly the saving on the spreadsheet doesn't look quite as clever. There's another option I don't hear discussed nearly as much.
Remote Australia.
Wollongong. Newcastle. North Queensland. WA.
You might get a different cost base without adding another country, another time zone and another layer of complexity.
I'm not saying don't offshore.
I'm saying do the whole sum first.
What We’d Watch Next
Whether more businesses start comparing offshore teams with regional Australian hiring before automatically assuming overseas means cheaper.
Are we paying for AI skills, or just expecting everyone to have them?
The headline salary market is fairly calm. Underneath it, the gap between general capability and scarce capability is getting wider.
Salaries aren't doing anything particularly dramatic overall.
Advertised salaries grew 3.5% in the 12 months to June.
But that number hides quite a lot. We're still seeing a premium for people with genuinely scarce AI capability. At the same time, I'm hearing this from employers:
“AI is just part of the job now.”
Which is interesting. Because if AI becomes an expected skill rather than a specialist skill, do you pay people more for it? I'm not convinced everyone has worked that bit out yet.
At the other end, the genuinely good people are expensive. And I think that's to be expected.
If one excellent engineer can do what you previously needed several people to do, paying that person properly might be the cheaper option anyway.
Want to know where you / your team sit?
Use the salary tool below. Pick the discipline. Pick the role. Add the package. We'll show you how it compares with the tekFinder 2026 Salary Data.
Salary benchmark
Pick your discipline and title, then add your package to see where it sits against the tekFinder 2026 Salary Guide.
Stop interviewing AI engineers like it’s 2022.
One of the better hiring ideas we heard this month was also one of the simplest. Let people use the tools they’ll actually use at work.
The suggestion was:
Let them use Google. Let them use AI. Then watch how they use it.
Give them 30 or 40 minutes and say:
“Build as much as you can.”
Don’t just judge the finished thing.
Look at the prompts. Look at the questions. Look at how they break the work down. Look at how quickly they know when the tool is taking them in the wrong direction.
That’s much closer to the job.
I think that's a much better test than asking someone to sit in a room and pretend the tools they use every day don't exist.
Anyone can rehearse an answer. The useful signal is whether they can think while the tools are available.
Try This
In your next technical interview, let them use the tools they’ll actually use at work. Assess how they use them, not whether they can pretend they don’t exist.
AI makes experienced engineers more valuable. Not less.
The code is getting easier to generate. Judgement isn’t.
I heard a good way of describing this recently.
The engineering role isn't disappearing. It's going up a level.
I think that's right.
AI can write loads of code.
Fine.
Who decides whether it's good code?
Who thinks about the architecture?
Security?
Performance?
Whether you should be building the thing in the first place?
That's where experience starts to matter even more.
A junior engineer can produce rubbish.
AI means they can now produce rubbish much faster.
The same goes for this sudden explosion of “AI Engineers”.
Using an AI coding tool doesn't make you an AI Engineer.
There's a big difference between using AI to write software and actually engineering AI into a product.
Agents. Evals. Memory. Models. Guardrails. Humans in the loop.
Different job.
One Question for Your Team
Are you measuring AI adoption by how much code gets generated, or by whether better products are getting shipped?
Now is a terrible time for great engineers to stop building.
This is the one I’m quite happy to argue about.
If you're a Lead, Staff or Principal Engineer thinking:
“Maybe it's time I moved into management.”
I'd think quite hard about it.
Not forever.
Just now.
Teams are getting smaller.
Good engineers are getting more done.
AI is changing how software gets built at a ridiculous pace.
And the people who really understand this stuff are becoming incredibly useful.
We're already seeing hands-on technical people earning what you'd traditionally associate with engineering management.
So why rush off the tools?
Management will still be there in 2 years.
This particular shift in engineering won't happen twice.
I'd be tempted to stay close to it.
Build.
Learn.
See where this goes.
Then lead.
Tell Me I’m Wrong.
Genuinely. Hit reply. I'd like to hear the other side.
Kind regards,
Nick Shepherd | Managing Director
Got a different read on any of this?
Good. Hit reply. I'd genuinely love to hear your thoughts.
Scaling your team?
Looking for staff?
Trying to work out what you should be paying?
Or just want a second opinion on what you’re seeing in the market?
Hit us up for a no pressure chat.
Nick Shepherd
Sam Lawson
Basil Benoiton
6.4% goes further than you think.
Did you know 6.4% of our placement fees support small charities through tekFoundation’s skilled volunteering services?
This month, that support is helping Dress for Success NSW & ACT, a charity supporting women into employment across NSW through its Styling service and online Career Hub.
The Career Hub is being revamped this year with help from tech volunteers.
If tekFinder’s worked with you, your fees helped pay for that. Thank you.
Find out more about tekFoundation →