Jens Lapinski's Recent LinkedIn Posts

Procurement is one of the largest levers inside big companies, but it’s rarely discussed when people talk about AI. 💭 In the latest episode of the Cutting Edge AI, together with Robin Harbort, I spoke with Fabian Heinrich from Mercanis about what agentic AI looks like in this industry. We talk through very concrete problems procurement teams deal with every day: long contracts, complex pricing structures, fragmented systems, and where AI already helps in practice versus where human judgment still matters. If you want a realistic take on AI inside complex, conservative enterprises, this one’s worth your time. 🎧 Listen here:

Angel Invest Coaching #72 - Frame the options smartly instead of arguing your point.

Angel Invest Coaching #74 - How to minimize risk as a founder.

Founder mode + AI = 🚀🚀🚀

Angel Invest Coaching #69 - If you feel like you’re racing against time to build your product, you’re not alone.

Angel Invest Coaching #68 - Reminder not to take things personally: people have demons you know nothing about.

Angel Invest Coaching #75 - Manage and maximize the energy level in your company.

Angel Invest Coaching #70 - The rise of the middle ground.

We keep seeing agentic AI in financial services framed almost entirely around payments. When you look a bit closer, a lot of the interesting work is happening elsewhere. Together with Denny Nguyen, we spent time looking at how agentic systems are being used today across wealth advisory, loan origination, and insurance claims: places where processes are still slow, manual, and fragmented. 🔄 We pulled those observations together in a 𝗻𝗲𝘄 𝗧𝗲𝗰𝗵𝗻𝗼𝗹𝗼𝗴𝘆 𝗕𝗿𝗶𝗲𝗳𝗶𝗻𝗴 𝗖𝗲𝗻𝘁𝗲𝗿 𝗿𝗲𝗽𝗼𝗿𝘁. If you’re building in fintech or AI infrastructure, I’d be interested to hear where you’re seeing agents actually work in practice. Full report here, a short preview below: https://lnkd.in/dRGuxdXn ⬅️

Angel Invest Coaching #76 - Don‘t become the bottleneck, take your hats off fast enough.

Angel Invest Coaching #71 - Sometimes you do everything right, and it still doesn’t work.

Angel Invest Coaching #79 – Give fewer answers, ask more questions.

Terrific job at a rocket-ship company 😀 🚀 🚀

Angel Invest Coaching #67 - Don't say yes if you haven't understood something.

Angel Invest Coaching #73 - The opportunity risk.

Angel Invest Coaching #78 - Sometimes nothing happens for a while and then suddenly you just takeoff.

Angel Invest Coaching #77 - Simple is always better.

There was an exit window for VCs in 2021. Most missed it. Too focussed on investing. Too focussed on growth. Back in 2025, my prediction for 2026 and 2027 was that these years were going to be big IPO years. This now looks correct. The logic is simple. We have Trump until Jan 2029. This is a stable exit window. 2028 is a year for distributing ownership in public stocks to LPs (lock up period etc). This means every company that can go public, should go public in 26 or 27. Stable window, appetite for certain deals, ample time to distribute. If you don‘t go public in 26 or 27 you might have to wait until the 2030s to get liquid. That is a long time away. There is also the Taiwan issue that is looming. I think it is very hard to predict 2029 onwards. I think most VCs will miss the current window. Again. Thinking about investing. And thinking about growth. And this time, their LPs will not just be annoyed. They will be angry. Good luck with raising that 2029 fund… Or maybe you make it rain and take your best assets public. I think that is the much better choice.

We just rolled out GRASWALD AI 2.0 — a major step forward in how fashion imagery gets produced at scale 🚀 Over the past few months, we've rebuilt the platform to become an operational layer for fashion content production, not just an image generation tool. Because content production teams don’t think in batches or experiments. They think in products, launches, and deadlines. With 2.0, work runs through projects— each product becomes a trackable task, and imagery moves through clear stages: Generate → Review → Approve So teams get complete visibility & control across PDPs and campaigns. The goal is simple: ✅ fewer regenerations, ✅ more control, ✅ and a workflow that matches how fashion brands actually operate. We’re onboarding new customers weekly — and we’ll share more as the rollout continues. 👀 👇 Here’s a sneak peek at the level of customisation & control the tool offers — in an easy-to-use, intuitive UI.

In our latest Cutting Edge AI episode, Robin Harbort and I spoke with Daniel Keinrath, co-founder and CEO of fonio.ai AI, about voice AI for SMEs. What I found most interesting is a simple point: in AI, the technology underneath is increasingly the same for everyone. So the real difference is not the model, but whether you build the product that actually works for the customer. That is also why the SME market may be more interesting than many think. Large companies can always decide to build things themselves. Smaller companies usually cannot. Listen here 👉 https://lnkd.in/dBEi5eXb Curious to hear how others think about this. Thanks to Daniel Keinrath for sharing his thoughts with us.

I recently joined the Question Everything podcast with Joseph Giuseppe Sofo and Jashar Seyfi. We talked about what I did before Angel Invest, but more importantly about a few things that come up again and again when working with founders. Most startups fail for fairly predictable reasons. For example: - Founder conflicts often appear around 6–9 months in. - Choosing your co-founder is one of the most important decisions you’ll make. - And many founders don’t spend enough time doing due diligence on investors. We also talked about the European venture ecosystem, what is holding it back, and why many of the problems are structural rather than cultural. Thanks to Joseph and Jashar for the invitation! 🎙️: https://lnkd.in/djMUmU2p

New episode of Cutting Edge AI with Ty Dunn, co-founder of Continue, together with Robin Harbort! One thing that seems increasingly clear: as AI starts writing more code, the role of developers shifts. Less time writing every line, more time orchestrating systems that produce, check and improve code. What that “mission control” layer looks like is still emerging. But ideas like continuous AI and automated agent checks are starting to point in that direction. 💭 We explored this and more in the episode. Full episode: