Denver, Colorado, May 20, 2026 —  I spent roughly ten years of my career working on the frontlines of humanitarian crises — with NGOs including Doctors Without Border and the International Rescue Committee, where I work now -  in places like Yemen, Syria, Haiti, the Congo, South Sudan, Pakistan, West Africa during the Ebola epidemic. It was high stakes work. The kind where the decisions you make and your performance on the job have soberingly real consequences for people, for whole communities.

But I’ll spare you the frontline stories — I’m going to tell you a story about helping people that was, for me, a way crazier journey.

After years on the frontlines, I found myself drawn to technology. I loved the excitement and optimism. Imagining a brighter future or new way to solve a problem. I took a few jobs in some crazy startups, and eventually a former colleague from the IRC reached out to me to lead a technology project that somehow combined both worlds. A project called Signpost.

I said yes. That was the best decision I ever made in my career. My name is Andre Heller and this is the story of our journey with Signpost (and you’re all a part of this story too…).

Signpost started in 2015. A small team. A simple idea.

Millions of people were on the move — refugees fleeing conflict in Syria, Afghanistan, Iraq, some crossing the Mediterranean toward Greece and Italy. And while aid agencies tooled up to deliver food, shelter, and medicine, everyone was struggling with something more basic: people didn’t know where to go, what their rights were, whether they could work, claim asylum, or rejoin a family member in another country. Or how to navigate the complex bureaucracies standing between them and those rights.

Information. That was the gap. Not just any information — information that spoke specifically to the problems they were facing in a rapidly changing context. Policies change. Services change. You get the point.

And here’s the thing that struck me about the aid sector: information was never treated as a priority. It was the afterthought. The assumption was that if you delivered the physical things — the food, the tents, the medical care — the information would sort itself out.

It doesn’t. Information is the glue that holds a crisis response together. Without it, people can’t access the services that exist. They make dangerous decisions based on rumors. They lose trust in the institutions trying to help them. They could even fall into the hands of human traffickers.

In a crisis, the right information at the right moment can be as important as anything else.

Signpost was built on two things: listening carefully before we act, and building trust by following our clients’ lead. We’d meet people digitally where they already were — on Facebook, on WhatsApp — and listen before we spoke. What did they actually need to know? What problems were they facing? What were they being told that was wrong? What did they trust?

So we created original content — accurate, vetted, up to date, user-centered — that spoke to the actual lived reality people faced in these crises.

Our mission: transform information into empowerment. Trust is the currency that makes it work.

By 2020, the project was alive — but barely. Life support at best.

Money was running out. The tech stack was a mess — expensive, fragmented, completely unscalable. We’d proven the concept but hadn’t figured out how to grow it.

I took the reins the month before COVID lockdowns started in the US. We’d already made a decision: kill it or scale it. Six months, maybe less. Either we committed fully — rebuilt from the ground up, found the right partners, went all in — or we shut it down gracefully and moved on.

And then the pandemic hit.

Face-to-face work — the default of the aid sector — wasn’t possible. Suddenly the thing we’d been building — reaching vulnerable people through digital channels — was exactly what the world needed. The opportunity was clear and we took our shot.

And that’s when we met Zendesk.

Sam Kind — where are you Sam?! — gave us his full attention, his expertise, and his genuine belief that what we were trying to do mattered. We rebuilt our entire technical foundation on Zendesk’s platform. For the first time, we had something that could actually scale. There were plenty of other tools we could have gotten for free — but this partnership was deeper than a deal.

The right partner isn’t the one who gives you a good deal. It’s the one who shows up like it matters to them too.

What followed was the kind of growth we didn’t dare dream of.

Six countries became ten, then twenty, then thirty. Afghanistan, Ukraine, Colombia, Mexico, Nigeria, Iraq – the list goes on. We had a dream team, a scalable blueprint, and we made bold decisions. We became a tech startup within the aid sector — a leading force in applied technology.

Three employees (myself included) became 10, became over three hundred in a matter of years. Twenty-five languages. Thirty countries. Twenty million users of our information products. 

In these short years – we did five hundred thousand digital social work consultations like this. That’s five hundred thousand people who sat down with one of our staff — digitally, one on one — and got the help they needed to navigate some of the hardest moments of their lives. So what does empowerment look like?

A family displaced by armed conflict in Burkina Faso is empowered to find shelter and access to the malnutrition program for their youngest child in desperate need of medical care.

A young man survives a difficult journey from Haiti to the Dominican Republic amidst gang conflict – he finds means to attain legal status and asylum and eventually bring his wife and children to safety. 

A 53 year old woman in El Salvador, a survivor of gender based violence, find programs and multifaceted support to help her through a transition towards a new life and recovery. 

In Kenya, a young refugee who was previously at a university pursuing his dream of journalism – all shattered by armed conflict - learns about a scholarship opportunity and wins it – he now runs a radio program for refugees.

A Kurdish refugee, facing persecution in Iran flees to Greece, and finds shelter, food assistance, and registers for asylum and a work permit (and later we wound up hiring him!)

20 Million users – and 500,000 one to one counselling sessions. Each one represents a story like this. All delivered on a shoestring budget.

That is our service model.

By 2023, the world noticed. We won the UN Sustainable Development Goal Impact Award, then the first ever and only European Union Humanitarian Innovation Prize. We took that recognition — and the funding — and made a bet: transform Signpost project – from a sort of scrappy SAAS model into, basically an AI company in the middle of the aid sector.

Let me back up on that. The AI transformation started earlier — December 2022. I’m in my car, driving, NPR on the radio. The story is about ChatGPT. Students using it to cheat on homework. I’m half-listening until I’m not.

I pulled over.

Because I knew — in that moment, with complete certainty — that our business model was dead. Not immediately. But it was in the mail. Everything Signpost did — creating content, answering questions, being the bridge between people in crisis and the information they needed — a language model would eventually do all of it faster, at infinite scale – and most likely do it at WAY lower quality, but it’d be cheaper. We had to get ahead of it – become masters of responsibly using it – and we were sitting on exactly the kind of original content and user interaction data that could make it really perform.

I went back to the office and told my team: Signpost is dead. Not the mission — the business model. We have no choice but to reinvent completely, or get bulldozed. Either we paddle out very early to catch this wave, or it’ll crush us once it grows.

In 2023, we secured a seat in Google’s first AI accelerator for nonprofits. Signpost AI’s strategy was not only to learn how to use AI — but to write the playbook for responsible deployment. Practical guidance, not just documents about principles. So while we prototyped and piloted, we published constantly — research, case studies, frameworks.

We were testing agentic AI in some of the highest-stakes environments in the world. The frameworks we built to orchestrate agents — and more importantly, the knowledge and know-how we built as a team — were useful not just for Signpost programs, but for the broader organization: refugee resettlement, education in displacement settings, combating human trafficking – the list goes on.

We were way out in the lead. It felt like a totally different startup. A real startup all over again.

Our new service model became empowering other teams to responsibly leverage AI towards diverse programming goals – and we totally reconfigured ourselves. 

And then came 2025.

Two things happened at once — and together they changed everything.

First: USAID was dismantled. European aid budgets redirected toward defense. The financial architecture that had held this sector together for decades was crumbling in real time – and continues to... Lifelines were cut en masse. Projects that had seemed secure suddenly weren’t. Signpost included.

Second: AI didn’t just get better — it got easier. No longer experimental. No longer just promising. Actually useful — reliable, scalable, in ways that changed what was possible almost overnight. What had required deep expertise in 2024 now required almost nothing. It’s not that our niche was in question – but the demand for our support on other projects started outpacing our capacity and people wanted AI everywhere – and Signpost was the team that did it.

We’d spent years as an independent project being strategic and sharp – and we were way out ahead.

But it turns out that as we rode the AI wave, it didn’t deposit us on the shore. It carried us into an entirely different ocean.

Sometimes you can see the destination on the horizon. Sometimes the ground moves under your feet. This time it was both — and we found ourselves paddling out again, but no longer alone. This time, everyone was in the water with us.

So against the backdrop of decreasing funds and pressure to leverage AI to pursue the highest priority work - we made the tough decision to scrap our independence and consolidate our tech team with another and to build something new — this time to support not only Signpost programs, but to bring AI to the organization’s highest priority work. It’s the right call. Resources need to go where they save the most lives. And it means that for me – personally – I have to let go of the team that I built – that’s the hardest thing a founder can do.

With this – it was essentially the big acquisition of our independent project, our little startup – and marks a new phase for the Signpost project and IRC – the end of one story and the beginning of a new one.  

And what comes next — starting now, in 2026 — isn’t about experiments with AI anymore. It’s about scale. Real, measurable, undeniable impact. For this sector, the bottom line isn’t just a margin — it’s the number of lives we touch. And AI gives us a chance to change what that number can be. And the IRC strives to meet this moment – this incredibly difficult moment for the aid sector and the communities that need us – and AI is probably our biggest bet.

And so through these transitions and shocks and massive changes – here’s what I actually believe, after all of it.

Our success wasn’t thanks to strategic thinking and bold attitudes. Not really. All of this matters of course. But the thing that made it possible, in every chapter, was people.

The team that said: six months, everything or nothing. The partner at Zendesk who showed up like it mattered to him. The three hundred people across thirty countries who showed up every day to do the real work. And the refugees who joined the team. The colleagues who stayed curious and kept driving forward even when the ground was shifting under all of us.

We’re in a truly transformational moment now and we have no choice about whether or not to embrace the future ahead of us.  It’s going to happen one way or another…

What we can choose — what has always mattered more than strategy — is who we choose navigate with. Not just people who share your goals. Goals change, strategies pivot, business models die and get reborn. Choose the people who share your values and conviction. Those are your fellow travelers.

I cannot emphasize this enough – we didn’t chose this historical moment that we’re in right now – these are crazy times for all of us.

You can really make a difference and you make a bigger difference than you realise.

To mine – the ones who were there through every chapter of this, who are still showing up now – to my fellow travelers at Zendesk, what a journey – thank you.