12 March 2025

Keeping Up With Edventures

Keeping Up With Edventures - February 2025

feb.jpeg

Excerpts from the GEM 2024/25 Global Report

Get the latest updates about our progress straight to your inbox by making sure you sign up for our newsletter, and be among the first to test our prototype by joining our Earlybird Community!

Dear Edventures community,

February was a month of both progress and setbacks and ultimately a test for both determination and motivation. It also helped us gain more focus and make harder priorities. On the product side, we improved coach insights, enhanced AI performance, and made usability updates, while also starting laying the groundwork for a more scalable Beta version. For the Alpha we’ll focus on maintenance and pilot-driven improvements, as we gradually shift our focus towards the Beta, needed for long-term platform growth.

That said, we have a great deal of updates to share with you, so happy reading!

If you’re just getting to know us, here’s a recap

Whether it’s your first or hundredth newsletter, I’m so happy you’re reading this newsletter and that you want to follow us on our journey towards becoming the global leader in entrepreneurship support, democratising entrepreneurship education and learning for all. You’re receiving this newsletter because we’ve been in touch for the past weeks, months or maybe years, and because your interest, skills, experience and/or professional focus aligns with our mission and values.

For those who are new subscribers, or to give you a quick recap, Edventures is an edtech and future-of-work startup based out of Sweden but with a fully remote and international team, spanning across three continents. Our mission is to develop a cutting-edge conversational AI that provides personalised coaching and learning experiences tailored to each entrepreneur’s specific business idea, needs, and challenges. Our platform also supports entrepreneurship organisations, coaches, and educators who play pivotal roles in supporting aspiring entrepreneurs.

Read on to discover this month’s latest updates and milestones. If you’re in a rush, below are the highlights you need to know about from last month.


TL;DR

  • Onboarded initial users for the JA Philippines pilot, and we’re already receiving feedback. Two main suggestions: easier BMC creation/editing and enabling coaches to access Anna too.

  • In terms of product development, we’ve revamped the coach insights and KPIs UI, improved AI performance with smarter prompts, smoother behaviour, and cost-saving updates, and started planning for a Beta version.

  • We were featured in the GEM Global Report 2024/25, the world’s leading study of entrepreneurship. Find our article on page 73 in the report.

  • In early February, we were approached by RISE about an EU project offering free supercomputer power to startups and academics.

  • The IEOx WinterChallenge took place between 1-15 February and was a great experience. We’re impressed to see so many well-presented and thought-through presentations. Congratulations to the winners: The Lambert Economic Team from the USA, runners-up The Sentinels from Vietnam, and third-place finishers The Keynesian Krew from Bangladesh!

  • Applications submitted to:

  • Despite setbacks with Eureka Innowwide, Social Tides, Qatar Investment Programme, and South Park Commons, we’re more determined than ever and preparing stronger applications for the future.

  • We’ve organised ourselves a bit more and updated the roadmaps for Alpha and Beta versions: Alpha focuses on maintenance and pilot-driven updates, while Beta is aimed at building a more efficient platform that is better built for long-term scalability.


February: Multiple Gut Punches but We Keep Progressing

⚙️ Product Evolution

February saw a lot of progress on both the front and back end. Since our latest update, we revamped the user interface for coach insights as well as its backend logic and endpoints to include 30+ different KPIs. This makes it easier for coaches to track their entrepreneurs, super admins to oversee coaches, and organisations to monitor overall performance.

Anna, our multi-agent AI, keeps getting smarter. We’ve iterated and fine-tuned system prompts, which improved both accuracy and response times. Along with this, we’ve implemented cached prompts, which temporarily store frequently used system prompts before sending them to the LLM API. This reduces and shortens repeated API requests, lowering API costs for us and, more importantly, for our clients. A real win-win!

Our improvements to system prompts now ensure a more natural flow between the different AI agents, whether chatting generally about business needs or more specifically improving a Business Model Canvas (BMC). Another major improvement is the separation of chat histories, an improvement only noticed in the backend. To the end user, the chat looks the same. However, now there’s also a separate version of this chat used for AI-driven analysis. This allows us to summarise conversations, conduct sentiment analysis, and provide coaches with deeper insights into what entrepreneurs discuss with Anna. This will lead to clearer coaching insights and more personalised guidance from Anna.

We’ve also made usability improvements. Users can now give direct feedback on Anna’s responses with thumbs-up and thumbs-down buttons. We’ve also added timestamps to the chat history for better tracking.

On the backend, we’ve implemented the first version of a CI/CD pipeline to automate and speed up deployments. Alongside these improvements, we’ve sorted out more bugs. One major bug was one where instructions sent to Anna weren’t properly reflected in the BMC. We also fixed a bug that sometimes jumbled the chat history in a strange (and quite funny) order.

We’ve started laying the groundwork for our Beta version by mapping updated user flows for entrepreneurs and coaches and creating low-fidelity wireframes.

The Alpha, built in iterations over the past two years by different developers, has served us well, but rather than patching existing flaws, we’re taking everything we’ve learned to build a lighter, faster, and more flexible Beta. This fresh approach allows us to integrate new features more efficiently and build an application that is more long-lasting.

Our long-term goal is to launch native apps with better localised user experiences, and optimising the codebase is a key step toward that. This transition won’t happen overnight, so while we develop the Beta, we’ll continue onboarding users to the Alpha and running pilots throughout 2025 to gather insights that will shape the next version.

Through our ongoing pilot with JA Philippines, we’ve received valuable feedback. One key request was an alternative way to update and complete a BMC, not only through conversing with Anna, which we fully agree should be in the platform and have now added to our Alpha roadmap. This update should be ready for our next update in March.

We’ve initiated conversations with NyföretagarCentrum in Sweden, both at the national and local levels, as well as followed up with JA Malta, JA Africa, and JA Singapore, along with some other organisations, as we in parallel explore more potential pilots.

🗞️ Edventures in the Media

On the 18th, the Global Entrepreneurship Monitor (GEM), the world’s leading study of entrepreneurship, released its 2024/25 Global Report. Alongside very interesting insights into the global state of entrepreneurship, Edventures was featured on page 73. Naturally, I’m thrilled to see us included, but more importantly, I’m excited to share our mission with a global audience that aligns so closely with our purpose. Have a read here!

Without getting too deep into the findings of the report, I’ve tried to keep the following analysis as brief as possible, yet contextually relevant. I’ll elaborate more on the findings in a separate blog post.

This year’s report highlights several key findings, particularly in areas closest to our mission—entrepreneurship education and support. Over the years, I’ve closely followed the GEM reports, paying special attention to three Entrepreneurial Framework Conditions (EFCs): Government Entrepreneurial Programmes (availability of high-quality support programmes), Entrepreneurial Education at School (early exposure to entrepreneurship), and Entrepreneurial Education Post-School (university-level training and incubation support). These categories are assessed by at least 36 national experts per country, ranked from 0 (very inadequate) to 10 (very adequate).

Once again, the findings confirm a troubling reality: entrepreneurial education is falling short. Across 56 economies, school-level entrepreneurship education scored an average of just 3.34, while post-school training fared slightly better at 4.81—but still well below an ideal level. This lack of early exposure leaves students unprepared to develop the mindset, skills, and risk tolerance needed to pursue entrepreneurship later in life. The problem isn’t new, but the data reaffirms just how widespread it is.

Across 56 economies, school-level entrepreneurship education scored an average of just 3.34, while post-school training fared slightly better at 4.81—but still well below an ideal level. This lack of early exposure leaves students unprepared to develop the mindset, skills, and risk tolerance needed to pursue entrepreneurship later in life.

It’s also clear that access to training programmes, incubators, and university courses remains inconsistent, and quality varies. Government support also lags behind, with an average score of 4.63 for entrepreneurship programmes. In fact, 56% of ranked economies fell below this average, and three out of five economies scored lower than the average for school-level entrepreneurship education.

The data is evident: without stronger education and better government support, we’re setting future entrepreneurs up for failure before they even begin. At Edventures, we recognise this problem, and we’re committed to changing it.

Without structured learning from an early stage, aspiring entrepreneurs enter the market unprepared, increasing their chances of failure—or worse, never starting at all. I’ve observed this first-hand through research and conversations with entrepreneurs over the years, and it’s precisely this gap that we aim to bridge with the Edventures platform.

February also brought an exciting collaboration with the IEOx WinterChallenge, where we provided this year’s business case. The competition received nearly 200 business case presentations, and I’ve started going through them one by one—my reading for the rest of the month is set!

If you’re curious about the finalists, you can watch the finals here and the closing ceremony here. Congratulations to the winners, The Lambert Economic Team from the USA, as well as the runners-up, The Sentinels from Vietnam, and third-place finishers, The Keynesian Krew from Bangladesh!

💰 Funding Opportunities

Not every month brings good news, and February came with its share of funding disappointments. Early in the month, we received the first setback from Eureka Innowwide. This time, I genuinely felt our application was much stronger than last year’s—we had a more relevant project partner, a more advanced platform, a stronger team and traction. However, when we reviewed the evaluation, we found systemic errors and inconsistencies. We requested a redress, but unfortunately, it was denied, officially closing the door on this opportunity. That said, Innowwide opens another call this summer, and you can bet we’ll be back with an even stronger application.

More recently, we didn’t make it into the third cohort of Social Tides’ GROW AI accelerator programme. On paper, we seemed like the perfect fit: a Europe-based startup with a strong social impact focus, AI at our core, an MVP in place, and great potential to scale. We ticked all these boxes, so the rejection stung a bit extra, especially since the timing felt ideal for that kind of support and cash injection. But, as with Innowwide, we’ll take the learnings from this experience and come back with an even stronger application.

Aside from these applications, we heard back from the Qatar Investment Programme, as well as from South Park Commons without a positive response.

On the brighter side, we were approached by RISE, Sweden’s research institute and innovation partner, about an EU-wide project offering startups access to free supercomputer resources. This is an exciting opportunity that could help us generate more training data for our AI and, ultimately, take steps toward training our own model in the future.

What’s in the Pipeline for Next Month?

  • In March we’ll continue to pursue more pilot projects, building on the conversations held in February, with the hope they will materialise more in March.

  • We’re updating our user analytics tool, Mixpanel, to gain deeper insights into how users navigate and interact with our platform. This will help us enhance the user experience in the Alpha, and refine future development of the Beta.

  • We’ll also act on the user feedback by making it easier for users to manually edit their BMCs, which is currently only managed by the AI. We’re also exploring the ability to generate BMCs by uploading documents to our platform.

  • We’ll be implementing several automations, both user-facing and backend, including onboarding and password reset emails, email verification upon signup, and auto-renewal of certain certificates like SSL.

  • As always, additional development will take place beyond what’s listed here—we continuously adapt based on new insights, feedback, and opportunities that emerge along the way.

  • We’ll be submitting applications for the Morgan Stanley Inclusive Venture Lab, the Civic Innovation Fund, and Norrsken Accelerator, and possible more.


Get the latest updates about our progress straight to your inbox by making sure you sign up for our newsletter, and be among the first to test our prototype by joining our Earlybird Community!


On that note, I want to wrap this month’s newsletter by thanking you for your continued support, and I look forward to sharing more updates with you next mont

Stay foolish, stay ambitious!

Alexander and the Edventures team