Corporate Venture Capital is the ultimate platform VC

From funding milestones to real-world traction: case study of our investment in Openlayer

The 'Why we invested' blog post is a VC staple. It’s full of polish, growth charts, and optimism, but for a 10-person Silicon Valley startup trying to sell into a ~10,000-employee European Telco, the 'Why' doesn't matter nearly as much as the 'How’. It’s always super interesting seeing impressive deal announcements on the internet and reading through the ‘why we invested in this company’ and the amazing growth plans of inspiring founders that we as VCs are backing, if we are lucky. For the founding team this funding announcement is obviously a great milestone (which shouldn’t go unnoticed), and definitely be marketed the best way possible to sign new customers, and attract the best talent. Our portfolio company ElevenLabs has been very tactical and unique the way they go about this, but that is for another post

However, it’s just the beginning. Companies can start accelerating and putting the fresh new funding efficiently to work to further grow and deliver on their usually (over) ambitious business plans. Similar to the founding team, as a venture capitalist this is when the journey begins. It’s the more exciting stage, throughout which we stay close to the founders and explore how we can support their growth, so an ‘over-ambitious’ plan turns into achieved results.

At KPN Ventures, the corporate venture fund of the leading telco in the Netherlands, our main strategic goal is to support commercial growth for our portfolio companies, while improving KPN’s product portfolio and operational excellence. So naturally, this is where most of our work goes into post-investment. Therefore, rather than publishing a traditional ‘Why we invested in Openlayer’ blog post, we wanted to share a memo reflecting on a successful partnership with Openlayer and looking ahead to the road still to come. We believe corporate venture could be the ultimate platform VC, if executed well. Our work with Openlayer will showcase that in more detail.

Solving the AI reliability problem at scale

Enough has been written about the ChatGPT moment in 2022, but this transformative moment has meant an inflection point where AI technology shifted from niche research to mainstream applications. We noticed early that as a telco, we’re not extremely well positioned to build frontier AI models given the high capex need, specific R&D, and engineering capabilities demanded to develop and compete at the frontier.

However, what we could do is secure and monitor these vulnerable AI systems. Similar to the way telco’s have offered secure connectivity solutions to customers to support them stay resilient in a digitalized world. We saw an opportunity to help both KPN and our large enterprise customers in deploying AI safely. Especially, given that this technology is still rather early in its lifecycle and facing many issues going into production safely at enterprise scale.

The goal for our participation in Openlayer’s Series A round, was making sure we can rapidly scale up our AI adoption at KPN, while remaining compliant for regulatory requirements such as the EU AI Act and NIS2. For every company, but especially a large regulated telco such as KPN, deploying the technology safely and responsibly is crucial. You don’t want to end up having hallucinating chatbots, AI voice agents giving away free data plans, or sharing sensitive customer- or company data right? That’s where the Openlayer team came in, led by co-founder and CEO Gabriel Bayomi. They are solving the AI reliability problem with their end-to-end governance and evaluation platform for agentic applications. A crucial and complex problem to solve, as reliability is harder for agents than for simple chatbots. An agent that can actually take actions in a telco billing system needs 10x the governance.

From design partner to distribution engine: getting Openlayer into the wild

So how hard can it be to sign large European telco customers with ~10,000 employees as a 10-person startup from Silicon Valley? Pretty difficult: long sales cycles, many compliance hurdles, several layers of business stakeholders, and unclear who is holding the ultimate budget. Especially complex for an AI governance startup that is adding value for multiple buyer personas such as product managers, software developers and AI governance leads.

We had seen and evaluated the Openlayer product during our due diligence together with our business teams, but bringing it from ‘could be strategically interesting for us’ to ‘we need this as our core governance and evals platform within our AI Excellence center’ takes time and effort. This can be hard to navigate with a small startup team and limited commercial track record within large enterprises. Helping them creating the right internal buy-in during a market scan, can accelerate the process for both the startup and the corporate here. In addition, for a large corporate such as KPN, engaging with fast-growing AI startups can also bring some unforeseen dynamics and issues compared to dealing with incumbent software providers.

Therefore we always try to stay as close as possible to both our portfolio companies as well as our business teams to make sure the partnership runs smoothly and any potential hurdles can be unblocked. Over the past year, we have really enjoyed seeing how the Openlayer team has been enabling KPN to further build out AI use cases, and get them to production safely. It really felt like a team effort where collective knowledge sharing was the key driver of success.

As mentioned before enterprise sales is super hard. We tried to support the Openlayer team in navigating internal processes, approving budgets and being a design partner by giving raw, honest customer feedback that a normal VC can’t provide. Luckily enough the business team fell in love with the platform quite quickly after testing it, which made our work a bit easier. We clearly remember Gabriel calling out one of our developers using the Openlayer platform for uncovering highly relevant product feedback. Insights that later helped Openlayer expand more easily to other European telco customers we introduced them to. It’s great to see how much value Openlayer has provided for our business at KPN and we truly believe more customers are going to follow soon.

“Collaborating with Openlayer was a true win-win that accelerated our GenAI vision. By sharing our internal research on evaluation metrics and helping them build a full AI Gateway integration, the platform quickly felt tailor-made for our eval-driven philosophy. This let us seamlessly scale our established way of working, while also enriching Openlayer’s out-of-the-box features and enterprise-readiness for their wider customer base” - Martin Kirilov, AI Tech Lead at KPN.

After a successful pilot across multiple business units, Openlayer is now live organization-wide, drastically accelerating our time-to-deployment. Teams can seamlessly swap LLMs, review tests in minutes, and confidently push to production while instantly catching anomalies, like silent provider updates or random 'goblin' hallucinations. Over the coming year, we will fully close the loop by making these specific tests and hard guardrails a mandatory part of our AI governance process.

Several large B2B customers in the Netherlands have already approached us about AI Safety and Governance, and we see strong potential for Openlayer and KPN to help our customers solve the AI reliability problem. We are looking forward to rolling-out more use cases, welcoming the Openlayer team back at our annual KPN AI Conference this summer, and debating football with Gabriel who is a huge PSV fan.

Leveraging the cap table for hyper-growth

The era of CVC as a 'strategic risk' is over. For the next generation of AI giants, the ElevenLabs’ and Openlayers’ of the world, the corporate investor is the ultimate 'Trojan Horse’: not a threat to the startup, but a trusted way into large enterprises. We don't just provide the capital; we provide the keys and navigation to the castle.

It’s exciting to see the growing role for corporate VCs as distribution partners and how sentiment is changing positively around the value add. Historically there used to be a rather negative sentiment among founders and operators, as there were potential risks by engaging with CVCs such as:

◆ loss of strategic flexibility: the corporation may pressure the startup to focus on projects that benefit the parent company rather than the startup’s own market opportunity

◆ restricted exit opportunities: the CVC may prevent the startup from being acquired by a competitor, or even taking a stake to acquire it down the line

◆ intellectual property exposure: close partnership leading to the Trojan revealing internal knowledge enabling the corporate partner to build a similar, better funded solution

What we currently see in the market, is a clear shift, in which high-growth AI startups are building GTM machines and leveraging all channels possible. Reaching $100M or even $200M ARR within 1 or 2 years of launching your first product rarely goes through one distribution channel, or solely focusing on B2C versus B2B. The next generation founders that are focused on extreme velocity are tackling them all at the same time. It is no surprise for us that companies such as ElevenLabs, Decagon, Legora, or Lovable, are all actively embracing corporate venture funds of leading companies globally to achieve their hyper growth state with long-term commercial partnerships.

In this world, the corporate investor (CVC) is no longer seen as a Trojan Horse threatening the startup, but rather as a positive Trojan Horse inside key markets, helping the startup gain trusted access to customers and distribution channels from within. And this is exactly our main goal at KPN Ventures: to help startups accelerate that flywheel and scale faster into the enterprise market.

So if you are building at high velocity and want to open up a strong distribution channel in the Dutch market, give us a shout! We are always keen on meeting ambitious founders, that are interested in Trojan Horses.

Daan Stolwerk, Investment Manager KPN Ventures

“KPN and KPN Ventures have been instrumental for our growth the past 12 months, and we really appreciate them as long-term partners to our business. Having them as a design partner, and seeing the platform live in action delivering actual customer results at scale has been great. Also their Venture team has made so many relevant warm introductions to their business colleagues, European telco friends, potential investors, and portfolio company founders, which has been extremely helpful.” Gabriel Bayomi, co-founder and CEO at Openlayer.

“In our day-to-day work, it’s sometimes difficult to keep track and staying on top of every latest technology or best platform in the market. That’s why it’s so helpful for us to have regular touchpoints with the KPN Ventures team as they help us find and navigate the best technology partners in the market”, Winifred Andriessen, Head of AI Excellence Center at KPN.

Martin Kirilov, AI Tech Lead of AI Excellence Center at KPN adds to that, “We are really happy to partner with Gabriel and the Openlayer team; their platform forms a strong foundation for building AI responsibly at KPN in the coming years. When you are constantly focused on building, taking a moment to connect with external experts and founders is incredibly valuable. That is why we are very grateful for our close relationship with KPN Ventures”.

‍