# Edtech Startup imagi Secures $4.5M to Bridge AI and Classrooms Safely > Female-founded edtech imagi raises $4.5M with support from OpenAI and Lovable to integrate AI into classrooms safely. > By Dave · 2026-07-24 > Source: https://otf-kit.dev/blog/imagi-ai-classroom-safety The K-12 edtech market has dropped 82% since 2021. imagi just closed an oversubscribed $4.5M seed anyway. That's the headline, but the interesting number isn't the dollar figure — it's who's on the cap table. OpenAI and Lovable staff are writing personal checks into the same company whose products they ship to classrooms. When the labs behind your distribution channel also bet on your roadmap, you're past "partnership" and into coordination. That alignment is rarer than the funding. ## Why safe AI in classrooms is a category, not a feature Most AI tools in education bolt safety on after the fact — a content filter here, a usage dashboard there. The harder problem is upstream: how do you design a product where the child is the primary user, the teacher is the safety reviewer, and the model is a black box? imagi was started in 2018 by Dora Palfi and Beatrice Ionascu, who met at KTH Royal Institute of Technology in Stockholm. They started as an all-female team working on why teenage girls often lose interest in coding around age fourteen — well before "AI safety in schools" was a funded category. Their first product was `imagiCharm`, a wearable device. The roadmap wasn't reverse-engineered from a 2023 model release; the company was already there. That timeline matters because the regulatory direction is treating educational AI as high-risk by default. The EU AI Act and most national frameworks put minors in the protected class. imagi's positioning as a "safety layer between advanced AI tools and schools" maps to that regulatory shape — and the founders have a head start on the compliance work that newer entrants will have to rush. Palfi put it bluntly: *"Today's schoolchildren are the first generation who will never know a world without AI. AI is not designed with children in mind. We need safe, tested AI education for children, with guardrails trusted by educators and parents."* That's the framing investors are buying. It's also the framing a district CIO will recognize from their own risk register. ![investor list, partnership list, and the OpenAI + Lovable staff overlap where capital and ](https://cdn.otf-kit.dev/blog/imagi-ai-classroom-safety/inline-1.png) ## What the $4.5M actually bought The round was led by `Brighteye Ventures` and `Day One Capital`. Other institutional investors: `FemInvest` and `We Are Human` — the fund started by Kahoot's founders. The angel list is where the signal lives: - `Mati Staniszewski` — ElevenLabs founder - `Thomas Dohmke` — former GitHub CEO - `Fredrik Hjelm` — Voi CEO - `will.i.am`, `Sophia Amoruso`, `Simon Squibb` - Executives from `OpenAI`, `Spotify`, and `Lovable` Three things to read off that list. First, the OpenAI + Lovable execs are double-dipping — they're distribution partners AND investors. That alignment matters when the product roadmap has to move at the pace of the underlying models. Second, the ElevenLabs and GitHub CEO checks hint at imagi's surface area being broader than "chatbots in class" — audio, code, and physical hardware are all in the product mix. Third, consumer-brand operators (will.i.am, Amoruso, Squibb) on a cap table tell you this company is being built as a brand, not just an enterprise sale. The consumer brand read is non-trivial. Kahoot built a $3B-ish edtech business by being the thing kids actually want to open. The imagi team includes people who have shipped hardware (Voi scooters), consumer audio (ElevenLabs), and developer platforms (GitHub). That blend is unusual for an 8-year-old seed-stage company. ## How to actually use imagi today This is the part most coverage skips. If you're an educator or a school IT lead reading about this round, here's what to do this week: ```bash # 1. Pull up the educator onboarding flow on imagi's site # The product is teacher-controlled, so the entry point is a # teacher dashboard, not a student app. # 2. Identify one classroom use case where AI replaces a worksheet, # not supplements a teacher. imagi's pitch is best-fit for the # kind of structured, repeatable task where a model can grade # or generate content and a teacher can review. # 3. Check your district's AI policy before piloting. The product # has guardrails, but the policy gate is yours, not theirs. # 4. Treat the OpenAI + Lovable integrations as a stability signal, # not a feature. If your underlying model provider changes # pricing or terms, imagi is the abstraction that lets you swap # without re-rolling the lesson plans. ``` The `imagiCharm` wearable is the on-ramp for the gender-gap-in-coding mission — a physical device students keep and use outside the classroom. The classroom tooling lives on web. The student-facing experience has to render the same way on a Chromebook, a school iPad, and the paired companion app. That's a multi-surface product, and the surface list isn't going to shrink. ## The female-founder signal in a contracting market K-12 edtech funding is down 82% since 2021. imagi closing an oversubscribed seed in that market, with a female CTO and a CEO who publicly cites the gender gap in coding as the founding problem, is one data point against the broader "female founders get less VC" narrative. But it's only one data point. The structure of the round matters more than the headline. imagi took money from `FemInvest` and from operators who have shipped consumer brands to the exact demographic — Kahoot for engagement, Voi for mobility, ElevenLabs for audio. That's a network, not just a check. A female founder raising in a contracting market on a mission-led thesis is newsworthy; a female founder raising in that market with this specific syndicate is a template other founders can copy. ## What this looks like under the tool churn Here's where the durable layer matters. `imagiCharm` is a physical device. The classroom tooling is web. The student-facing experience has to render the same way on a Chromebook, a school iPad, and a paired phone. The underlying AI providers — OpenAI today, others tomorrow — will change pricing, deprecate models, and shift capabilities every quarter. The part that doesn't change is the cross-surface UI. The same component has to behave the same on web and on a paired mobile companion, with a token-aware render path so a Chromebook on a throttled school-WiFi connection still loads. That's the durable layer: shared primitives that survive the model churn. If you're a team building this kind of multi-surface product — physical device plus web plus mobile, with a model layer that's a moving target — that's the bet worth making on infrastructure that doesn't rewrite when the model release notes drop. imagi's product team is going to spend the next 18 months on exactly this problem. ## The cap table is the moat The underrated takeaway: imagi's defensibility isn't the safety layer (commoditizing fast) or the brand (early). It's the investor list. When the AI labs themselves have skin in your roadmap, the partnership is structurally harder to walk away from than a vendor agreement. OpenAI has both reputational and financial reasons to keep `imagi` alive and well-integrated. For educators, this means `imagi` is more likely to still be around in three years than the average seed-stage edtech. For builders, the lesson is that in a category defined by safety and trust, the moat is relationships with the platforms you depend on — not the code. The kids losing interest in coding at 14 have been waiting since 2018. The $4.5M is just the receipt that someone, finally, is building for them.