For December, instead of taking a look back at recent news, we're looking into the future.

We sat down with Niel Robertson, founder of Winslow, to go under the hood of the most cutting-edge HRIS we’ve seen so far.

If you’ve wondered how HR software changes when AI isn’t just bolted on but is the entire operating principle, this one’s for you.

Why AI-native is different from AI-powered

Robertson: ““AI-powered today feels like there’s a button somewhere that slides out a window and there’s a chat embedded on top of the existing application.”

Most vendors today add a chat box on top of their app. It can summarize a policy or point you toward a menu, but the heavy lifting still happens in the old interface.

Niel’s framing is simple: AI-powered chatbots tell you how to do something. AI-native agents help you do it.

Ask an HR chatbot, “Do we have Presidents’ Day off?” and you get a clear answer.
Ask that same chatbot, “Add a new dependent,” and the system points you to a how-to guide.

This is where the AI-native difference shows up.

Instead of a trail of tabs, Winslow brings the task-specific widget right in the conversation so you can complete the change without leaving chat.

Context is a small thing, but it's everything

Robertson: “To be good at AI technology in HR, you have to be very good at context.”

AI that answers questions is easy. AI that answers your question, for your situation, is the hard part.

Winslow started by solving a very real HR headache: repetitive employee questions that require policy nuance and individual context.

If two employees ask the same question from different states or pay groups, the correct answer can diverge quickly.

Winslow’s approach is to encode company policies, locales, and individual attributes so the system is context-aware by default.

Getting the organizational context right unlocks trustworthy answers and safe actions, which is the foundation for everything else.

The system should build itself

Robertson: “You should assume employees are always just asking the AI how to accomplish things - and the system should extend itself to make that true.”

Traditional HRIS platforms try to be everything for everyone. The result is a sprawling interface that most employees barely touch.

Winslow flips that model. You get a strong core for payroll, benefits, people records, onboarding, and automation. Then you extend the platform with only what you need.

Two real examples from Niel’s demo:

• Parking pass request. Instead of spinning up a Notion page or spreadsheet, you describe the workflow once. The system instantly builds a mini-app that collects the right fields and routes approvals.
• T-shirt size collection. Rather than teaching 450 people how to navigate to a custom field, you enable “update my T-shirt size” and the system handles the form, storage, reminders, and reporting.

This is an HR system that evolves with you. You aren’t stuck buying every feature up front. You start with a lean core, then vibe code small, purpose-built pieces that fit your company like a glove.

The interface should be broken apart

Robertson: “Take any kind of SaaS application and break it into tiny little bits and pieces - and when that piece is the right answer to your question, just bring that piece into the chat, let someone interact with it, and go back to what they’re doing.”

The old SaaS model is one large desktop screen with infinite nested menus.

AI-native platforms, like Winslow, deconstruct the application so that you only see the relevant elements at the moment you need them.

You move fluidly between text answers and focused UI fragments. You never hunt for the right tab. The system simply surfaces the piece you need.

It also lives where your people already work.

The same capabilities can be exposed in Slack or even inside ChatGPT, so casual users never have to “log into the HRIS” at all.

User gradients, not one-size-fits-all

Robertson: “If employees never even see a SaaS application, they’ll be very happy. That’s the beauty of a conversational interface.”

Not everyone wants to, or needs to, be a power user of HR software. AI-native design respects that:

• Employees ask questions within the chat interface, complete tasks, and move on without touching a maze of menus.
• Leaders can query ChatGPT for insights without ever logging into the HRIS or learning the reporting function.
• HR Power Users still have a full-featured HRIS to dive deep into when they need to, but they are no longer the help desk for “where do I click?” questions.

The result is less retraining every open enrollment, fewer “Help me" emails, and more time for strategic work.

What this adds up to

The future of AI-native HRIS might look something like this:

You start with a lean core for payroll, benefits, records, and automation. Then you continuously 'vibe code' the additional functionality your company actually needs.

When employees ask for something, the system responds with the right form of help in the right place. Sometimes that’s a direct answer. Sometimes it’s a focused widget that captures data, triggers approvals, or kicks off a workflow.

And it all travels to where your people already work, whether that’s the HRIS, Slack, or ChatGPT.

That’s the promise of AI-native: a system that understands your organization, surfaces just the piece you need, and quietly builds itself around the way you actually work.