Friends,

Do you feel like you’re on top of things when it comes AI?

For every 10 People professionals I speak to, 9 of them seem to say they feel behind. So this edition, I thought I’d gauge the readership

(Please vote! I’ll share results in next weeks edition).

And while this won’t be a solid indication of actual capability, it will at least be a gauge on sentiment.

Part of me wonders if most of our sense of being behind is just the plethora of people saying they’ve built ‘$10k per month businesses with AI’ and they do it by selling you a course (which ironically is based on telling others you make passive income and selling them a course… rhymes with pyramid scheme, oh wait no thats just what it’s called).

Anyway, this week we’re looking at a team that truly is at the cutting edge of AI use (at least compared to what I see in the market). The cool thing even if you do feel behind in all of this, reading articles like this is what helps you further ahead!

Enjoy this week’s edition ✌️

Matt

In case you’re new here (or just missed it) here’s the past three editions of the FNDN Series:

Data safety and keeping pace topped the list of concerns. Job loss came in behind both.

Inside the report are more data points that put AI adoption into perspective:

The mid-year State of AI in the People Function report gives a clear picture of where AI has shifted how People teams work in the first half of 2026, where it stalled, and what leaders are watching next.

Thank you for supporting our sponsors, who keep this publication free.

Know a startup Head of People looking for answers 🙋 why not forward this to them for some instant karma? ✨

This edition is based on the latest episode of the FNDN Series podcast, with Hallie Condon.

Hallie Condon is VP of People at Chili Piper, a 150-person B2B SaaS company that helps revenue teams turn more leads into meetings and customers. She owns comp, performance, hiring, and AI adoption, and loves digging into the broken parts of HR that make people uncomfortable.

Of all the People teams I speak to, each one is at a different stage when it comes to the role AI is playing in scaling their team.

What is consistent, is that teams are telling me they’re increasingly being expected to do more with less.

In this piece we take a look at Chili Piper, with a 2 person HR team supporting a workforce of around 150 spread across 36 countries.

In my conversation with their VP of People, Hallie Condon, we discussed how the team started to scale its capability through AI, the advances they're taking it to, and the unanticipated impacts it's having on the People function and how they're trying to combat them.

If you don't know Marie Kondo, then clearly you were doing something way more fun in 2019 than me.

She became famous for her minimalist decluttering system, which instructs people to keep only items that "spark joy".

(Suffice to say, I got rid of a lot of sh*t that year)

And it's this concept we touched on as a method for identifying the kinds of things People teams should use to determine what work AI should handle, and what it shouldn't.

If you're an avid reader, you'll already know I spoke to former guest Ben Langner about the things AI shouldn't automate (and how to build your no-fly list), so the 'things that bring you joy' is almost like a complementary layer to this.

Because when the opposite is true, and you're giving AI all the things you enjoy and left with the admin that drains you, you've made your own job worse. It’s something I've caught myself doing, where I’ve outsourced work in a way that doesn't make me happy, and handed over parts of my role I actually love. 

Bleh.

So how did Chili Piper tackle this?

One thing that didn’t bring them joy, was checking the annual learning and development budget and how it was being applied.

Their capacity (i.e. only two people) limited their ability to effectively check each claim against the policy, and meant they were only able to do spot audits. 

Using AI, they were able to build the process and controls that would assess expense claims against their policy amount, preventing inadvertent overspend and saving their time through automation.

And presto, they eliminated something they didn’t enjoy (checking expense claims). PLUS it was able to scale their team from checking some claims, to all claims! Of course their finance team would love them too because now they had less risk of leaky expenses, and the process was being applied more fairly.

Side bar: Another thing that comes up around judgement with AI, is how you’re actually using it. AI can help you build a deterministic process, or a probabilistic one, and it’s important to choose the one that’s right for the scenario.

For example; if the policy is 'approve all L&D expenses below $500', a claim that comes in for $450 is yes, and one that comes in for $550 is no. It's an 'if this then that' type of arrangement.

Probabilistic problem solving (i.e. what AI does) would look at that question, and the answer amounts to a roll of the dice because it's predicting a likely answer, rather than running the rules.

This is the same issue as the 'how many r's in strawberry' that AI keeps getting wrong.

So if we know something is low joy, and relatively low/deterministic judgement, we can use that criteria to easily categorise other things that we can to evaluate AI for in the People team.

Create a two by two matrix, one axis with joy, and one with judgement.

If something is low joy and low judgement (i.e. relatively deterministic), it's ripe for automation.

Be weary of the work categorised ‘low joy / high judgement’, because the temptation is to hand it over to automation when it’s often the work (involving difficult conversations) that nobody wants, but absolutely requires a human in the loop.

High joy and low judgement is the trap I described above, the work you love that happens to be easy to automate. Let’s make sure not to give that up too easily.

And high joy with high judgement is the work all of this is ultimately buying you time for.

What impressed me next was where things are going, and I think remote companies have a distinct advantage here.

I’ve worked in proper remote companies before that actually had dedicated headcount and resources around how teams worked in that environment, so I was excited to hear about this next stage of the conversation and just where this concept was going.

In Chili Piper's example, they already had a strong ‘documentation’ muscle. For example, company decisions were made based on documented business cases. Amazon calls them PRFAQ’s, others might have different names, but the premise is the same; you document key strategic decisions and people can review it and refine it before you get started.

In Chili Piper's case, capturing key moments of company context was something they did well already, but were looking to take further.

What they built is a company brain (and it’s increasingly being discussed as a future state of AI in companies).

A company brain is a set of institutionalised knowledge that people access and contains the entire company's context, the decisions that have been made, the goals, and (in Chili Piper’s case) a library of skills anyone can run.

Hallie planned to launch version one with ten skills. She shipped a hundred and twenty, and I loved hearing about the real world implications.

We all know what someone's onboarding costs a company. Working out who matters, what good looks like in the role, where things live, and who to ask about what.

Traditionally that comes out of your team's hours, one conversation at a time, and in the current environment it’s increasingly time people don’t have to give.

So Chili Piper built role packs.

Every role gets a context document covering its goals and its key stakeholders, with the skills relevant to that role sitting alongside it. When someone onboards, the system surfaces three skills specific to their job and explains how each one helps them, then asks where their friction points are and recommends more based on the answers.

You might remember me mentioning company brains as a component of some companies adopting the ‘Player-Coach’ leadership model, and how it was predicated on the ability for a company brain to effectively perform the role of priorisation and direction setting for a team. I don’t know if we’re there yet, but it sounds like that’s where it’s going.

As a remote workplace, Chili Piper already had the ingredients they needed with their documentation culture, and AI gave them the means to turn it into something the whole company could draw upon.

I pointed out earlier why I think remote companies have the edge here: The documentation culture came first. AI only made it useful after. 

Companies should think cautiously about adopting a similar approach without that culture in place already, or risk institutionalising the wrong kind of decisions.

Which, when you're two people supporting 150, is the difference between the thing that saves you and the thing that buries you.

One of the things that showed up in our conversation about how far Chili Piper had come in scaling their team (and how far they were going) was actually what was being lost along the way.

When the HR team are traditionally the ones that people go to to solve some small part of their day, there's inevitably some other context that comes along with it, and that does a couple of things:

Both of these are lost when the person can self-serve their needs to the highest levels imaginable, and it's the loss that was pointed out as a side-effect of robust AI automation.

And being candid, Hallie didn't have a solution for it, and in fact she was grappling with the ethical conundrums of what it actually could be. One consideration was some kind of sentiment analysis across the interactions people were having with AI (the company brain), and Claude itself challenged the ethical implications of reporting on it.

Her instinct was that it risked feeling 'big brothery', which sits badly against how much Chili Piper cares about autonomy. 

The challenge that persists is how to maintain that efficiency in a way that doesn’t sacrifice the people in People & Culture.

The pattern in Chili Piper's journey wasn’t complicated (although some of the outcomes certainly were).

When it comes to scaling the team, people should start with work that's joyless and needs no judgement. 

Then you build the knowledge layer so decisions stop walking out the door with the people who made them, and give the whole company a way into it.

But the thing to keep an eye on in the meantime is what risks being lost out the other side. 

For every efficiency, are you giving up some part of the interaction with your people that you still need in order to be successful?

If you enjoyed this post or know someone who may find it useful, please share it with them and encourage them to subscribe.

A roundup of the most interesting stuff from the week:

When you’re ready, here’s three ways I can help you:

1. Tools & resources
Resources and tools that give you what you need to build your own startup compensation practices.

2. Comp consulting
Building startup compensation practices that are clear, fair and competitive.

3. Startup People Summit
A 1-day annual event for People professionals in scaling companies. Creating the playbook for startup people practices. Grab recordings from past events, or subscribe to join the next summit.