Primo's MCP server: Your AI assistant can now talk to your IT operations

The benefits of automation are well known to modern businesses. For decades, companies have found ways to turn slow, repetitive processes into efficient, self-executing systems. Which lets teams focus on impact, rather than repeating the same low-value tasks. 

Onboarding and offboarding are both high-value processes made up of low-impact touchpoints. Getting team members up to speed quickly really matters. How you create their email account or reset security permissions doesn’t.

Which is why automating these manual steps makes such a big difference. Automated onboarding and offboarding takes low-value work off your plate, and lets you focus on what is important. It also makes both processes faster, easier, and eliminates basic errors. 

In this article, we look at how automation can improve your onboarding and offboarding processes—particularly for IT operations. Then we meet two companies who successfully automated their own IT onboarding, and saw tangible benefits. 

‍

What are employee onboarding and offboarding processes?

Onboarding and offboarding are the practical, functional, and cultural processes associated with welcoming and farewelling company employees. Onboarding typically includes teaching new hires about the company culture, training in your specific ways of working, and giving them the hardware and software tools they need to execute. 

Offboarding is the change process at the end of an employee’s time with your company. This can include exit interviews, farewell celebrations, and regaining possession of company property like computers, phones, and access cards. 

‍

Key steps in IT onboarding

The IT onboarding process is often slower than you’d like. It involves numerous distinct steps, which can really add up if handled individually and manually. These include:

• Setting up user profiles and permissions
• Ordering new devices
• Configuring applications, software, and security updates on these devices
• Delivering devices to new employees
• Training employees on compliance, cybersecurity, and optimal use
• Monitoring device performance and troubleshooting issues

‍

IT is just one aspect of an employee’s onboarding, and can be taken for granted by hiring managers. Your goal is to make all of the above happen smoothly, quickly, and with no extra work for yourself or the new hire. 

For help, see our short checklist for efficient IT onboarding. 

‍

What IT offboarding involves

While the IT onboarding process may be neglected, offboarding is often overlooked altogether. Retrieving devices from departing employees is essential both for asset management and security. 

Key steps include: 

• Locking devices the moment employees no longer need them
• Wiping personal data or returning devices to factory settings
• Returning physical devices to the office or supplier
• Checking a device’s state for reuse
• Preparing devices to be redeployed

All of this adds up, and is always more complicated with remote or distributed teams. In a traditional office setting, it’s pretty simple to have an employee hand in their devices on their last day. It’s more challenging if that employee is in another city, state, or country. 

‍

Why automate employee onboarding and offboarding? 

In general, the best processes to automate involve a number of manual steps and little added value from having people handle each one.

Key benefits of automating your IT onboarding and offboarding include: 

• Time saved for IT teams and hiring managers, who no longer need to manually work through each of those steps we saw above. 

• Faster onboarding for new employees, who don’t need to wait for people to set up their profiles or order devices.

• Near-instant offboarding, because devices can be locked or wiped immediately with a simple click.

• Fewer errors, including skipped or forgotten steps, faulty devices, or losing track of devices when an employee leaves. 

• More consistent experiences, as every employee follows the same automated process at the beginning and end of employment. 

Overall, automation creates more streamlined and efficient internal processes. And for something as common and recurring as onboarding and offboarding, efficiency gains can really add up.

How modern SMBs automate onboarding and offboarding — and why it works

To illustrate with tangible examples, let’s take a look at two companies that prioritize automation in the onboarding and offboarding process. 

Like many growing companies, both faced real challenges in scaling IT operations. Even as modern tech companies, they had few resources specifically for IT operations. They needed to create efficient, easily-replicable processes to get new employees up and running, and to smoothly offboard team members at the end of their work. 

Best modern SMBs have understood that a great onboarding experience comes from the collaboration between HR and IT teams — and these two companies made that alignment a core part of their approach. As we’ll see, the secret to success lay in choosing the right tools and partners to take the weight off their very busy leaders.

‍

‍

Faume: Near-instant IT operations for a distributed workforce 

Founded in 2020, Faume is a technical logistics solution that lets brands create resale services for their products. Faume works with world-famous logos like Hugo Boss, The Kooples, Aigle, and Bash to bring second lives to items and make consumer commerce more sustainable.

Faume’s 30-person team includes remote staff across France. CTO and Co-founder Jocelyn Kerbouc’h needed a simple way to deploy and manage devices for this distributed workforce ahead of scaling post-Series A.

‍

Before: False starts with IT providers

Faume initially leased computers in the hopes of getting additional support and a streamlined service. But this was far more expensive than the cost of buying—they were asked to pay up to €2,500 for a €1,200 computer. And worse, they still regularly encountered malfunctioning devices and frustrating errors. 

They pivoted to buying from Apple directly, tracking devices manually in a Notion doc. This was certainly more cost effective, but added more administrative effort to the onboarding process. 

As a co-founder wearing multiple hats, Jocelyn couldn’t afford this extra admin. Faume needed a more robust IT operations solution that could deliver devices at the right price, while also tracking their use and ensuring security. 

‍

Today: Centralized IT onboarding & offboarding

The big switch was finding an IT operations provider that lets Jocelyn order, configure, and deliver employee devices in a few clicks. Using Primo, Jocelyn sets password rules and updates, and pre-configures applications so that computers arrive ready to use. 

“Thanks to Primo, onboarding new employees now takes us half the time it used to,” says Jocelyn.

Faume has essentially automated the onboarding process, and offboarding is just as simple. When an employee leaves, Jocelyn can lock and wipe their computer remotely. Departing employees receive a shipping box and can easily return computers from anywhere. 

The result is a more efficient, secure IT environment for Faume. And Jocelyn can put all his energy into building and leading his business.

Read the full Faume story here.

‍

‍

Dalma: Efficient operations with no IT team

Dalma is France’s fastest-growing pet health insurance company. Its tech-enabled platform already insures more than 40,000 European cats and dogs, with no signs of slowing down. 

Founded in 2021, the 70-strong team has grown quickly to deliver this popular and worthwhile service. While that’s good for business (and for our pets), it put pressure on former Head of People Claire Maarek. 

With IT onboarding just a small portion of her role, Claire didn’t have the time or technical expertise to build a comprehensive program from scratch.

‍

Before: Poor leasing experience

Like Faume, Dalma also tried leasing as a (theoretically) efficient way to manage IT operations. But Claire explains that the downsides were obvious right away. “Our leasing experience was disappointing, offering minimal service and reliability with poor customer support.”

It was a maddening mix of high prices and low-quality service. For an HR leader like Claire—not an IT pro by trade—this wasn’t a tenable situation. 

‍

Today: IT onboarding in seconds

Since switching to Primo, the results are night and day. IT onboarding takes mere seconds, and Dalma can secure hardware at competitive prices, configured and delivered for when the person arrives. All of this with no deep IT procurement knowledge or dedicated technical experts. 

Most importantly for HR professionals, Primo integrates with Payfit (alongside other HR platforms). Dalma adds a new employee in Payfit, and most of the process is automated from there. Devices arrive on time, whether new hires are in France or Germany. 

When an employee leaves, Primo makes it easy to retrieve or reassign devices elsewhere, or simply resell them. Which makes both onboarding and offboarding as easy as can be. 

Read the full Dalma story here.

‍

‍

Make IT onboarding and offboarding a breeze

Both IT onboarding and offboarding are relatively simple processes, made difficult by manual steps and a need for technical expertise. Particularly for growing companies without IT teams or paid external consultants, key steps can fall through the cracks. 

That’s how you end up with security risks, sluggish processes, and frustrated team members — right when first impressions matter most.

The best way to streamline IT onboarding and offboarding is with one central solution. And as both Faume and Dalma showed, it’s even better when that solution integrates with your HR systems and company tools. This lets HR leaders and hiring managers—often “accidental IT managers”—keep control and ensure each step is completed efficiently. 

Primo provides exactly that: an all-in-one IT management system for faster onboarding and offboarding. You can easily automate virtually all of your IT operations, without paying huge fees to managed providers. 

‍See how Primo can improve your IT operations today.

A new hire’s Day One says everything about how your company runs. If their laptop is on the table, configured, logged in, with the right apps installed — they feel set up. If the laptop is still in a courier’s warehouse and IT is scrambling to provision SSO — they feel like an afterthought, in their first six hours.

This is the IT onboarding checklist lean teams actually use. Five phases, RACI ownership, HRIS-triggered, mirrored for offboarding. Use it as your operating playbook, not a static doc.

The five phases, at a glance

1. Pre-boarding (weeks -4 to -1): order the device, create the IdP account, provision baseline SaaS by role, configure the device for delivery.
2. Day One: device delivery, SSO login, MFA enrollment, EDR check, Acceptable Use Policy signature.
3. Week One: verify every tool works, complete cybersecurity training, schedule first manager check-in.
4. First 30 days: audit installed apps against role profile, document additional access requests, validate everything still works.
5. Offboarding parity: set up the reverse workflow on Day One, not on exit day.

Each phase has owners, a timeline, and a measurable outcome. Done well, the whole flow runs in the background of an HRIS event. IT only gets pinged on exceptions.

Phase 1: Pre-boarding (weeks -4 to -1)

Most IT onboarding problems are pre-boarding problems. If the laptop is ordered late, Day One can’t be saved. If the IdP account is missing, the SSO chain breaks on first login.

HR-to-IT handoff (intake trigger)

The trigger should be the HRIS event, not a Slack message, not a calendar invite. The minimum payload IT needs to act:

• Full legal name and preferred name
• Start date and timezone
• Role and department
• Manager
• Work location and shipping address
• Hardware preference (if you offer choice)

If you’re capturing this in a Notion form or a Slack thread, you’re one departing teammate away from a gap. Make the HRIS the source of truth and pipe events into your IT workflow tool.

Order the device for direct-to-employee delivery

The window from order to first power-on is the most expensive part of IT onboarding when it goes wrong. Lead times for the EU vary by hardware and reseller; for international hires, add a week for customs.

The right pattern: order through a procurement workflow that ships the device pre-configured directly to the employee. If your platform supports zero-touch deployment, through Apple Business Manager or Windows Autopilot, the device finds your MDM automatically on first power-on, with no IT touch between order and Day One. Primo’s procurement workflow covers 60+ countries with delivery in around 5 business days, with apps and security pre-configured before shipping.

Create the user record in your IdP

Identity is the spine of everything that follows. The IdP account (Microsoft Entra ID, Google Workspace, JumpCloud, your federated IdP of choice) is what every SSO-enabled app will check against. Create it as soon as the HRIS event fires, not on Day One morning.

Provision baseline SaaS access by role

Group memberships in your IdP should map to roles, not individuals. “Designer” gets Figma, Notion, Slack, the design Drive. “Sales Rep” gets HubSpot, Gong, Slack, the sales Drive. Maintain the matrix once; reuse it for every hire.

Phase 2: Day One

A well-run Day One feels boring to IT and magical to the new hire. That’s the goal.

Device delivery and unboxing

If pre-boarding was done right, the new hire receives a sealed box, powers it on, connects to Wi-Fi, and watches the device configure itself. No IT presence required. This is the payoff of zero-touch deployment, and it’s the single most visible signal that your company runs operationally.

First login and SSO verification

The first login should be against your IdP. The new hire enters their company email, completes the IdP flow, and lands on a configured desktop. If they have to type a separate password into anything besides the IdP, your SSO chain has a gap. Fix it before Day One, not after.

MFA enrollment

Enroll the new hire into MFA during the first session. Use a phishing-resistant method (passkey, hardware key, or platform authenticator) wherever your IdP supports it. SMS-based MFA is below the line in 2026. Keep it as a fallback for account recovery only.

EDR agent and security policy check

Endpoint Detection and Response (EDR) should be installed by your MDM as part of the configuration push, not by the user. Verify in the admin console that the agent is reporting healthy before the new hire opens their first customer call. While you’re there, confirm disk encryption (FileVault, BitLocker), firewall, and idle-lock are all green.

Acceptable Use Policy signature

Push the AUP as part of the Day One flow, captured digitally with timestamp. Same for the phishing-awareness module assignment. This is the boring half of compliance, and the half that pays back during your next audit.

Phase 3: Week One and first 30 days

The first week is verification. The next 30 days is calibration.

In Week One, confirm every tool the new hire needs actually works: VPN, conferencing, email signature, calendar permissions, shared drive access, and the second-tier apps that came through the role profile. Schedule the first manager check-in for end-of-week-one, not later. Complete the cybersecurity training module.

In the first 30 days, audit the installed apps against the role profile and document every additional access request that came in. If a single role is generating 10+ ad-hoc access tickets in month one, the role profile is wrong, not the workflow. Fix the profile, not the ticket.

Use your HRIS as the trigger, not a spreadsheet

The single biggest upgrade you can make to IT onboarding isn’t a better checklist. It’s connecting your HRIS so the checklist runs itself.

When a new hire is created in an HRIS like BambooHR, HiBob, Factorial, Eurécia, Deel, Dayforce, Charlie, ADP or Gusto, the right remote device management platform should:

• Create the IdP account
• Order the device through the procurement workflow
• Pre-register the device to the OEM portal where supported
• Assign role-based MDM and SaaS policies
• Send the Day One welcome guide
• Alert IT only if something needs human intervention

This is the model behind Primo’s IAM page summary: “HR triggers it. Primo executes it.” Events flow from HR’s source of truth straight into device, identity and access workflows.

The win isn’t only speed. It’s parity: every new hire gets the same baseline, regardless of whether IT was busy that week.

Provision software and access by role, not app by app

If you’re provisioning SaaS access one app at a time, per hire, you’ve already lost the next ten hours.

The discipline that scales: define role profiles once, then map every new hire to a role. The role determines the apps, the permissions inside those apps, and the IdP groups they belong to.

Apply the Principle of Least Privilege as defined by NIST: each role gets only what’s needed to do the job. Run access reviews quarterly to catch role drift.

A minimal role-profile matrix:

• Designer: Figma, Notion, Slack, Drive. Editor on design assets, viewer elsewhere.
• Sales Rep: HubSpot, Gong, Slack, Drive. CRM rep view + own pipeline.
• Engineer: GitHub, Linear, Slack, Drive, AWS. Repo write on owned projects, AWS dev only.

Role profiles also make offboarding meaningful — you know exactly what to revoke, because you defined it once when you hired the role. Primo surfaces this directly as “Role-Based Access Control (RBAC) across every app”.

Procurement is part of onboarding

This is the part of IT onboarding most checklists skip, and most lean IT teams quietly burn weekend hours on.

A flawless Day One can’t recover from a laptop that arrives late, arrives unconfigured, or arrives at the wrong address. Procurement isn’t a separate vertical. It’s the first stage of onboarding.

For a lean IT team in 2026, procurement should cover:

• Sourcing through authorized resellers (so OEM zero-touch works where supported)
• International shipping with customs handled
• Pre-configuration before the box ships
• Asset tracking from purchase order through delivery
• Return labels generated automatically for the eventual offboarding

If your current setup is “IT lead orders devices manually, ships from home, types serial numbers into a spreadsheet” — that’s the part of the workflow with the highest return on automation. Primo’s procurement workflow handles this end-to-end: “From order to delivery, Primo ships, configures, and tracks every device automatically.”

Build one checklist for onboarding and offboarding

The most expensive part of offboarding isn’t the wipe. It’s the SaaS account nobody owned that quietly retains access for six months.

Every line on your onboarding checklist needs a mirror on your offboarding checklist. Build them at the same time, not on exit day.

The same HRIS event that started onboarding can fire offboarding. Primo runs this as “Zero forgotten access. Ever. From first day to last, every account, seat, and permission is managed automatically.” and “Revoked automatically on their last day to prevent security breach.”

Without that pattern, a wiped laptop doesn’t reclaim Slack, Google Workspace, HubSpot, or Stripe. Identity and device have to be revoked together — same workflow, same trigger.

Frequently asked questions

What should be on an IT onboarding checklist?

A complete IT onboarding checklist covers pre-boarding (hardware ordering, account creation in the HRIS and IdP, baseline SaaS provisioning), Day One (device delivery, SSO login, MFA setup, EDR install, acceptable use policy signature), and the first 30 days (training, tool verification, access audits). It should also establish offboarding parity from day one.

What is the difference between IT onboarding and HR onboarding?

HR onboarding covers contracts, payroll, benefits, culture and orientation. IT onboarding covers everything the new hire needs to work on Day One: hardware, accounts, applications, security setup and policies. In practice the two should be triggered from the same HRIS event so they stay in sync.

When should IT onboarding start?

IT onboarding should start at least two to four weeks before the new hire’s first day. That window covers hardware ordering and shipping, account creation in the IdP, baseline SaaS provisioning, and any zero-touch deployment configuration. For remote international hires, add another one to two weeks for customs and delivery.

What does a new hire need on Day One?

A configured laptop, working SSO login, MFA enrolled, email and chat access, calendar synced, role-based app access, an installed EDR agent, and a signed acceptable use policy. They also need a working manager check-in and a help channel for IT issues.

How do you onboard a remote employee?

Ship a pre-configured device using zero-touch deployment. Trigger account creation from the HRIS so credentials are ready on Day One. Provide a written Day One guide. Schedule a video onboarding call with IT and the manager. Verify SSO, VPN and MFA remotely. Set up a clear escalation channel for first-week issues.

How long should IT onboarding take?

Pre-boarding spans two to four weeks. Day One setup should take under an hour for the employee if zero-touch deployment is in place. The full onboarding cycle, including training, access audits and role validation, typically runs 30 days. Anything longer suggests manual handoffs in the IT-HR workflow.

Who is responsible for IT onboarding?

On lean teams responsibility is shared: HR or the hiring manager triggers the workflow, IT executes provisioning, and the manager validates role-specific access. A RACI matrix prevents gaps. On smaller teams without a dedicated IT person, an HR or office operations lead often owns the IT onboarding workflow.

See an HRIS-triggered onboarding flow that handles device, identity and access from one console, with offboarding parity built in.

Every day there's a new AI model. A new benchmark. A new funding round. A new outage. A new "this changes everything" thread. A new paper that supposedly makes the last paper obsolete.

I build and sell an agentic IT platform for a living. I'm supposed to be on top of this stuff. And honestly, I spend a non-trivial amount of my week just trying to keep up with all the AI news.

So I sat down and wrote this, partly to organize my own thinking, partly because I suspect I'm not the only one feeling this way. If you're a founder, an operator, or an IT leader trying to make real decisions in the middle of all this noise, I hope some of it will be useful.

Here's where I've landed on AI in 2026, and what I think it means for the future of agentic IT.

The age of abundant AI is ending

For most of the last three years, frontier AI felt like an infinite resource. You picked a provider, wired up an API, and assumed the best models would keep getting better, cheaper, and more available.

That era seems to be ending. And I don't think most companies have caught up to what that means.

Demand for frontier AI is outpacing supply, and the constraints are physical: energy costs, infrastructure bottlenecks, the economics of serving billions of inference requests per day.

• In April 2026, OpenAI paused its Stargate UK data centre project, citing energy costs and regulatory uncertainty. That project was meant to deliver 8,000 GPUs in Q1. It delivered none.
• Nearly 50% of planned data center projects in the US for 2026 are facing delays or cancellations.
• GPU rental prices for Nvidia's Blackwell chips have surged 48% in 60 days. • CoreWeave has raised prices 20% and extended minimum contracts from one year to three.
• OpenAI's CFO said on the record that the company is "making some very tough trades at the moment on things we're not pursuing because we don't have enough compute."• Anthropic has shifted Enterprise billing from flat per-seat fees to per-token pricing. The subsidies are ending.
• Anthropic has removed Claude Code from the Pro plans while admitting they’ve also made other small adjustments (e.g. weekly caps, tighter limits at peak), citing “usage has changed a lot and our current plans weren't built for this”.
• Starting June 1, 2026, GitHub is also shifting all GitHub Copilot plans to a usage-based billing model.

When supply is scarce, providers prioritize the customers who pay the most. The investor Tomasz Tunguz recently described five characteristics defining this new era:

1. Relationship-based selling (SOTA models reserved for strategic customers)
2. AI to the highest bidder (prohibitive pricing for everyone else)
3. Available but slow (no performance guarantees)
4. Inflationary commodity pricing (demand compounding against fixed supply)
5. Forced diversification (developers pushed toward smaller models, open source, or on-prem until infrastructure catches up)

The moment that made this all real for me was when, in April 2026, Anthropic released Claude Mythos Preview, which the company describes as a step change over its previous models. In internal testing, it autonomously discovered and exploited zero-day vulnerabilities in every major operating system and web browser, including a 27-year-old bug in OpenBSD. Normally a capability jump like that would kick off a months-long race between labs to ship their own version. Instead, Anthropic did something unusual: it chose not to release the model publicly at all.

Access to Mythos is reserved for a consortium called Project Glasswing. The members: AWS, Apple, Cisco, CrowdStrike, Google, JPMorgan Chase, Microsoft, NVIDIA, Palo Alto Networks, and the Linux Foundation. The US Treasury has publicly requested access and is expected to receive it. Anthropic is giving these partners $100 million in usage credits to find and patch vulnerabilities in critical infrastructure. Everyone else, including the vast majority of companies that have spent the last three years building products on Anthropic's API, does not get Mythos. It's the first time in nearly seven years that a leading AI lab has so publicly withheld a model from general availability.

There are legitimate safety reasons for that decision. I'm not criticizing it, though some will argue this is a PR stunt to reinforce Anthropic’s safety-focused positioning. But as a founder building on top of this infrastructure, I can't ignore what it can signal. If the strongest models move toward a world of consortium access, strategic partnerships, and government briefings, with only hyperscalers and bigger institutions getting a seat at the table, what does that leave startups and mid-sized companies?

If you're building on frontier models right now, this is the reality you're planning against, whether you realize it or not.

AI is more operationally fragile than people admit

The other thing I don't see founders and leaders talking about enough is how unreliable AI systems still are in production.

1. Amazon spent early 2026 dealing with this firsthand. Its internal coding agent Kiro autonomously deleted a production environment, causing a 13-hour AWS outage. By March, a string of AI-assisted code deployments took down Amazon's retail website for hours, locking millions of shoppers out of checkout and wiping an estimated 6.3 million orders in a single incident.
2. A rogue AI agent at Meta posted internal information that led an engineer to accidentally expose sensitive company and user data to unauthorized colleagues for two hours, classified internally as a "Sev 1." A Meta safety director publicly described how her own agent deleted her entire inbox, despite explicit instructions to confirm before taking any action.
3. /PocketOS, a company that sells software to car rental businesses, went into chaos mode after a rogue AI coding agent deleted the company’s entire production database and its backups.

Put the three together and you get a clearer picture of the operational reality underneath AI in 2026. The tech is transformative but it's also unstable in ways that matter if you're running production systems.

Some companies are already hedging

This is where it gets interesting for me, because some companies aren't waiting to find out how the scarcity and the model provider dependency stories play out. They're getting their hands dirty.

Intercom recently launched Fin Apex 1.0, a customer support model built on an open-weights foundation that the company says outperforms frontier models on its specific task. Their thesis is blunt: pre-training has become a commodity. The real edge is in post-training, fine-tuning models on your own domain data until they beat the generalists at the thing you actually care about.

Cursor, the AI code editor that hit $2 billion in annualized revenue in early 2026, built its Composer model on top of Kimi K2.5, an open-source model from Chinese lab Moonshot AI. That only came out when a developer intercepted Cursor's API traffic. It sparked a broader conversation: the most capable open foundations available today disproportionately come from Chinese labs. DeepSeek, Qwen, Kimi. These are the models companies are quietly building on when they want performance without frontier-lab dependency.

This is arguably the worst-kept secret in Silicon Valley. And it tells me something about where the market is heading: toward a world where application companies own more of their stack, and frontier labs become one input among several, not the input.

Where I think durable value actually gets built

Here's the part I've been thinking about the most, because it's the part that informs how we build Primo.

Hebbia's George Sivulka articulated it better than I could in a recent piece: foundation models, no matter how powerful, will never know how your specific team does its specific work. He calls it "process engineering." Software isn't just code. It's a stored process. It encodes the way a specific team cooperates on a specific problem. The private credit desk at one firm uses different compliance flags than the private equity team at the same firm. Two IT managers at the same company will have entirely different standards for how onboarding should run, how access reviews happen, how tickets get triaged.

Foundation models can't be opinionated about any of that because they're built for every use case on Earth at once. They can't know, and frankly don't need to know, the specific preferences of any particular team.

That's the opening for vertical software because the institutional knowledge encoded inside is what’s valuable.

What most people get wrong is that better foundation models don't erode vertical software. When reasoning models like OpenAI's o-series shipped, everyone predicted legal AI would get crushed but the opposite happened. Vertical legal AI had its best year ever, because stronger models made the orchestration layer more reliable, not less. The orchestration layer is where the trust lives. You can have the most capable model on Earth and still produce garbage outputs if you don't have the scaffolding to constrain, verify, and route that capability through a specific professional workflow.

2025 was the year AI became truly useful for law. 2026 is becoming that year for finance and cybersecurity. I'd argue it's also that year for IT.

What this all means for agentic IT

So here's where I land, after talking to many IT teams and stepping back from all the noise.

How do you build something durable in a world where your underlying models may become gated, expensive, slower, or quietly different from the ones you shipped with?

Below are a few ideas that I often come to and that shape how I think about Primo and IT.

The moat is the process. Your IT team's workflows, your company's compliance posture, your specific onboarding and offboarding flows, etc. None of that lives in a foundation model. It lives in the software that encodes how your team actually operates. That's the layer that gets more valuable as models improve, not less.

Agentic IT is empowerment. I say this a lot, but I mean it more in 2026 than I did a year ago. The compute constraints and the rogue agents examples aren't arguments against AI. They're arguments for keeping humans in the loop where the stakes justify it. The IT teams that win with AI aren't the ones that hand everything over and hope for the best. They're the ones that use AI to handle the routine stuff so they can focus on the work that requires judgment, relationships, and context no model will ever have.

You need an AI-native stack. You cannot place AI agents on an IT stack that doesn’t have the right context, knowledge, and the proper data to work with. The underlying stack that AI communicates with is critical for quality answers and actions. In IT, you need to own the APIs, you need quality workflows, a system of record, and strong integration with HR systems.

The short version, for anyone who skipped to the end

AI is real. Agentic IT is real. But the infrastructure underneath both is more fragile, more political, and more economically strained than the marketing suggests.

The companies that will win the next few years aren't the ones with the best AI demo. They're the ones that understand their moat isn't the model, it's the process knowledge they encode into the layer on top of it. They're the ones that build agentic systems as extensions of their IT teams and can do more with the same resources.

AI doesn't need you to believe in it uncritically. It just needs you to use it well.

If you're overwhelmed by the pace of all this, you're not alone. I am too. But the signal underneath the noise is clearer than it looks. I hope this helped uncover it a little.

‍