Global tech giants are lining up to work with a 25-year-old Kiwi, and the profile fits Auckland startup Atomic Tessellator. Founder Alain Richardt’s company uses AI to design substitutes for rare earth elements, compressing materials discovery that normally takes years. This is exactly the high-value, IP-heavy export story New Zealand keeps saying it wants. It is also exactly the kind of company New Zealand has a habit of losing.
A problem big enough to bring in the CIA’s money
Rare earths sit inside permanent magnets, defence systems and data centre hardware, and China dominates the supply chain. A credible way to engineer around that dependency is not a science fair project. It is strategic infrastructure.
The investor list says as much. Atomic Tessellator’s seed round was led by London’s Crane Venture Partners, with In-Q-Tel, the CIA-founded venture vehicle, alongside Icehouse Ventures and Outset Ventures, whose investment committee includes Rocket Lab founder Sir Peter Beck. Selected for the Google for Start-ups Accelerator while working on a samarium alternative, the team received US$350,000 in compute credits and ended the programme simulating 680,000 atoms, modelling materials at nanometre scale. Amazon’s AWS programme added US$100,000 in credits.
That is a Kiwi founder solving a bottleneck the world’s largest companies would happily pay to remove. It deserves to be celebrated without qualification.
The talent is not the problem
New Zealand does not have a founder shortage. It has a retention problem. Kiwi founders sit behind more than ten billion-dollar companies built almost entirely offshore, including Anaplan, Supabase, Wayve and Substack. NZGCP calls each one a “double loss”: the country misses the original value creation and the flywheel of alumni and capital that turned Xero and Trade Me veterans into the next wave of local founders and angels.
The domestic picture is lopsided too. About 77% of the tech sector’s enterprise value sits in just three firms. Halter, the latest unicorn, is now incorporated in Delaware and run from Colorado. New Zealand’s success rate moving from early-stage to major rounds is 23.7%, below the OECD average of 26.7%.
The mechanism is no mystery. A study of roughly 11,000 venture-backed startups found about 6% relocated internationally but they accounted for 17% of value created, with foreign, particularly US, capital strongly linked to the move. Atomic Tessellator already has a London lead investor and a US intelligence-linked backer. The gravitational pull is obvious.
Going global is fine. Leaving everything behind is not
Nobody sensible wants to stop a deep-tech company chasing customers in Washington or Silicon Valley. A five-million-person market cannot sustain that ambition alone. The right question is what stays here.
Professor Rod McNaughton of the Centre for Innovation and Entrepreneurship puts the venture-backed sector at an estimated NZ$133 billion, but argues the headline number matters less than whether founders, staff and investors recycle knowledge and capital into the next ventures.
There is proof it can work. Christchurch AI firm Partly raised US$50 million in a Series B led by DST Global, opened a Texas office, and still plans to add at least 60 roles, mostly in Christchurch, which remains its R&D hub. It is even pulling Americans and Europeans south. Hnry has similarly stayed headquartered here while expanding into Australia and the UK.
The numbers that should worry Wellington
Tech is already a heavyweight. In 2024 the sector contributed $23.8 billion to GDP and $11.4 billion in exports, third behind dairy and tourism. But the R&D base underneath it is wobbling. Business R&D spend rose just 0.4% to $4.1 billion in 2025, R&D intensity slipped from 0.98% to 0.95% of GDP, and full-time R&D staff fell 5.1%.
That is the context for any founder deciding where the lab, the patents and the senior engineers end up. Grants have not changed the pattern. What would is deeper local growth capital, so the Series B and C rounds do not automatically come with a US redomicile clause, and tax and visa settings that make it rational to keep R&D teams onshore.
What happens next
Atomic Tessellator will need far bigger rounds to turn simulations into commercial materials, and the biggest cheques will almost certainly be foreign. That is not a failure. The test is whether, five years from now, the modelling team is still in Auckland, the IP is still owned here, and Richardt’s early employees are writing angel cheques for the next deep-tech founder. Partly shows it can be done. Halter shows how easily it is not. For a country that wants exporters built on brains rather than bulk commodities, this is the one to get right.
Sources
- Stuff: Why 25-year-old Kiwi has global tech giants lining up to work with him (2026-10-03)
- NZ Herald: Auckland’s Atomic Tessellator uses AI to synthesise alternatives to rare earth elements. It could recast geopolitics (2026-06-25)
- NZGCP: New Zealand Tech Ecosystem Report 2026 (2026)
- Newsroom: Invest in startups’ IP to keep them here, or watch our unicorns gallop offshore (2026-07-21)
- The Conversation: NZ creates remarkably valuable tech firms. How can it keep more of the benefits? (2026-08-16)
- Centre for Innovation and Entrepreneurship: New Zealand’s venture-backed tech companies are now worth an estimated NZ$133 billion (2026-08-20)
- NZ Herald: New Zealand’s AI start-up, Partly, is taking on the world from Christchurch (2026-09-10)
- Tech New Zealand: Aotearoa New Zealand’s Tech Ecosystem (2025-09)
- Stats NZ: Research and development survey 2025 (2026-04-23)
Join the discussion
Add useful context, ask a good question, or challenge an idea — keep it specific and respectful.