Not an Update. A New Way Forward:
The People Behind Zammo Agents

Handing the Tools Back to the People Who Know Citizen Service

Ask anyone on the team what they want people to take away from this launch, and a theme shows up again and again: this wasn't built to turn government staff into AI experts. It was built to hand the tools back to the people who already understand citizen service best. 

Stacey Kyler, our Managing Director, put it bluntly. A lot of AI products quietly ask your team to become technologists just to use them. Zammo Agents was built the opposite way: staff build and improve agents themselves, in plain language, no dev cycle required. 

"Agentic AI doesn't have to mean giving up control. Done right, it hands more control back to the people closest to the problem, while still letting IT have oversight of the software." 

That same instinct shaped the product itself. For Patrick Gilligan, our Director of Product, what makes Zammo Agents a new product, not just an update, comes down to the architecture. The whole thing was built around the outcome you want to achieve, not a flowchart of every possible path. Instead of manually mapping every scenario and edge case, you tell the agent what you're trying to achieve, and it works through the rest. 

That shift shows up fast in what a customer can actually do. Patrick's example is a deliberately unglamorous one: filing an IT ticket. In the old world, covering what happens when the system's down, or a resident changes their mind mid-conversation, meant hours of manual scenario-mapping. Now, you describe the brand and persona you want, and the agent handles those fallback cases out of the box, testing itself as it goes. 

"What used to take real thought and a lot of scaffolding, you can now do in a matter of minutes." 

The feature Patrick's most excited for customers to try is the one behind all of it: the Agent Designer. You describe what you're trying to build in plain language, and the system asks the follow-up questions, fills in the gaps, and assembles something powerful, without you ever needing to know what's happening underneath. 

None of that erases what's genuinely hard about building for government, and here, Patrick doesn't point to the AI at all. He points to compliance. SOC 2, accessibility, procurement checklists: government customers sit at "the tip of the spear" on all of it, and every one of those boxes has to be checked before the technology matters at all. 

For Patrick, the simplest way to sum up Zammo Agents skips the technical pitch entirely: it's a one-stop shop, like ordering off a dinner menu. You say what you want, and the system figures out how to build it, deploy it, and keep it running.  

To be clear, this is a new product entirely, not a replacement for the original Zammo platform. If you're already running UI Builder and it's working well for you, it isn't going anywhere. Zammo Agents is there for teams who want impact without the technical flow building. 

The Engineering Behind the Ease

None of that simplicity happens by accident. For James Mann, our Platform Engineer, the toughest technical challenge wasn't the AI at all. It was the customers already depending on the old product, who couldn't afford to be disrupted. 

"We had existing customers, and they couldn't be impacted at all. We had to work out how to get to where we ended up without breaking it for them." 

That same instinct, reliability over flash, shapes how James thinks about "no-code" as an idea. Plenty of AI products let you describe an agent in a paragraph of plain English. James is candid about why that's not enough on its own: prompt-built agents drift. Someone tweaks a line of text, and the behavior shifts in ways nobody can predict or guarantee. 

So instead of hoping a prompt holds up, the team built guardrails in as structure, not suggestion, like requiring a customer to explicitly opt an agent into connecting with their systems, rather than leaving that judgment to the AI. And there's a trust benefit non-technical users notice immediately: connect the agent to real data, and you can watch it get built in front of you. 

"That's what gives you the confidence it's actually going to work." 

The part he's proudest of is one most customers will not notice, until they need it: a feature called Agent Log, giving full visibility into not just what a resident asked and how the agent responded, but why the agent did what it did. Customers had been asking for that kind of transparency for years; the old architecture made it genuinely hard to deliver. This time, it was built in from day one. 

Boiled down to three words, James describes the platform as intelligent, autonomous, explainable, and he's specific about why the third one matters as much as the first two. 

"The more an agent does on its own, the more you need to see what it's doing to actually trust it. Autonomy without explainability is scary." 

It's a familiar irony in good engineering: the more invisible the work, the harder it usually was to build. 

Meeting Skepticism Head-On

Not every conversation about AI in government starts from excitement. Some start from real, reasonable hesitation. 

Alex Farr, Zammo CEO, says his answer to agencies on the fence isn't to argue the fear away. It's to validate it, then redirect it toward action: 

"It's okay to be concerned about AI, but what is not okay is to write it off as something that will not be needed." 

The pain point heard most, before an agency ever talks to Zammo, is almost always some version of: AI is expensive, scary, and we don't have the technical depth to manage it. Alex's response leans into specifics over reassurance, solving three out of four citizen queries; paying for itself in under five months; running 24x7x365 across languages and channels. And when agencies want proof before they commit, the team can build a demo on their own data in under 48 hours. 

"We're not about preoccupying you. We're about helping you take action." 

What Success Looks Like From Here

For Stacey, success isn't a timeline, it's a picture of how things actually feel once this is working. 

For agencies, she says, success is almost boring: a resident calls in, gets an answer, hangs up without ever thinking about the AI behind it. Staff spend their day on the conversations that actually need a human, not the fortieth "what are your hours" call. 

"If agencies are quietly running more of their citizen service through Zammo Agents because it's just working, that's success." 

For Zammo, success looks like becoming the platform agencies point to when they describe what "good" AI in government looks like, not just the fastest to build with, but the one they trust most. 

And if she had to describe Zammo Agents to a city manager who's never touched agentic AI? No jargon, just the resident's side of it: 

"Picture every question your residents already ask you, permits, bills, trash pickup, and imagine all of it just getting answered instantly, no matter how they reach out. That's Zammo Agents. And the part that tends to surprise people is that your own team builds it, not developers, not a six-month project." 

The Heart of It

Strip away the ROI figures and the platform architecture, and this launch comes down to something simpler: a group of people who believe government service should feel as easy as everything else in a resident's day, and who built the thing that makes that true, for the people who already do this work every day. 

That's Zammo Agents. Not a feature update. A different starting point. 

Ready to see it for yourself? Schedule a no-cost, custom demo scoped to your agency's use case or try it for free first and then reach out when you've seen what it can do.