- Sofia-based Thinkpilot raises EUR 600K pre-seed from LAUNCHub Ventures to build its AI product development co-pilot
- The team tackles chaotic product ideation by integrating context-aware AI into workflows
- The platform offers personalized, structured support for product teams’ decision-making
- The investment strengthens the team, refines the product, and accelerates market engagement
This May, Thinkpilot—the Bulgarian AI startup developing a co-pilot for product teams—annouced having raised its Pre-Seed round of investment. The amount is EUR 600K, and the investor is the well-known Bulgarian VC fund LAUNCHub Ventures (invested in Lace AI recently, among others).
Founders with Firsthand Experience Solving the ‘Messy Middle’ of Product Development
A diverse team of co-founders started Thinkpilot in late 2024 in response to the frustration of product teams struggling with scattered, chaotic information through their decision-making process. Having spoken with dozens of teams, we realized the problem wasn’t just messy data, it was the whole messy middle of product ideation, where data, strategy, and decision-making all collide.
Thinkpilot’s CEO Stoimen Veselinov comes from a SaaS product background and has an excellent idea of the pain of having to lead product discovery without clear data or structured input.
As an engineer, Kiril Dimitrov, the head of AI, has been building high-scale machine learning models and personally dealt with having to fill in gaps left by unclear product direction. This gave him an immediate understanding of how AI could solve this problem for product teams.
Antonio Iliev, the startup’s CTO, is a former lawyer turned senior engineer and team lead. In his experience leading engineering teams, he saw quite a few promising projects fail due to poor alignment. This motivated him deeply to help fix the disconnect between product and engineering.
Finally, Megan Wolff—Thinkpilot’s head of growth—originally joined to help with early go-to-market and messaging. Her strong background in product marketing and running product discovery processes in SaaS brought invaluable expertise in understanding how product teams think and operate, making her the perfect fit to help shape Thinkpilot.
Stoimen Veselinov, Co-Founder and CEO at Thinkpilot
‘None of us came into this chasing hype. We’ve each lived the pain of product misalignment in different ways. That’s what convinced us we were the right team to build Thinkpilot, a product built for the product builders we’ve been ourselves,’ Mr Veselinov tells ITKeyMedia.
The CEO admits that getting people interested turned out harder than the team had expected. They leaned on their network, particularly investors, who helped open doors to potential users and get the ball rolling. Then, all founders needed to be put together, aligned, and committed.
Product Discovery Challenge and Avoiding the ‘Build Trap’
‘Convincing each person wasn’t about selling a dream, it was about finding shared pain and a genuine passion for solving it. There were tough conversations about equity and roles. It felt risky,’ Mr Veselinov recalls.
Antonio Iliev, Co-Founder and CTO at ThinkPilot
‘On the product side, we had the vision, but not the solution. We had to iterate a lot. That’s emotionally draining. You get excited, then realize the idea doesn’t land. But what helped us was our mindset: we weren’t married to the solution, we were committed to solving the problem. That kept us going,’ Mr Iliev adds.
It does happen quite often that teams get together, build cool tech, and only then look for a problem to match it. With such an approach, teams end up hallucinating problems just to justify their tech stack. This is the build trap that Thinkpilot committed to avoiding at all costs. Instead, the team anchored itself in the real-life issues product teams face, and only then started thinking about how to solve them with tech.
‘One great example: most people would solve the “information overload” problem using RAG systems to surface documents. But that approach creates hallucinations and overloads people. We rejected it. Instead, we built a context-aware system that pre-filters relevant information per initiative, which creates less noise, better accuracy. It’s harder to build, but it’s the right solution for the problem,’ Mr Veselinov explains.
Tackling Product Teams’ Real Pain Points with Purpose-Built AI Workspace
According to ProductPlan’s 2023 report, cross-functional alignment confuses 47% of PMs, while 38% feel hindered by scattered feedback. In response, Thinkpilot facilitates structured, AI-assisted decision-making for product teams by providing context within their existing workflows.
Unlike tools that add AI upon legacy interfaces, Thinkpilot’s solution represents a dedicated AI workspace designed to support product teams throughout their workflow. It brings together a collaborative assistant and task-specific agents to help with ideation, research, validation, and specification, keeping decisions aligned with data, strategy, and user input.
Kiril Dimitrov, Co-Founder and Head of Machine Learning at ThinkPilot
More specifically, Thinkpilot’s platform focuses on the most common workflows across product teams for them to get started fast with minimal friction. Long-term, the AI copilot adapts to how a particular team works. Teams upload their strategy documents, the system ingests this context, and over time, our models will learn and tailor themselves to your specific processes. Mr Veselinov emphasizes that their platform isn’t a one-size-fits-all product but the groundwork for seamless personalization.
‘Thinkpilot integrates directly with tools like Jira, Confluence, Google Drive, and more, to pull in relevant company context continuously. It syncs tickets, documents, specs, and research to create a live product knowledge layer,’ Mr Dimitrov explains.
Thus, when shaping new initiatives, Thinkpilot automatically surfaces related work, past decisions, and supporting materials from these sources, reducing duplicate efforts and preventing missed dependencies. The AI co-pilot then uses this context to suggest next steps, highlight risks, and help product teams move from idea to structured requirements faster and with better alignment across stakeholders.
From Code to Product Decisions
Todor Breshkov, Founding Partner at LAUNCHub Ventures
With AI speeding up software development, the focus increasingly shifts from implementation to making informed decisions about what to build and why.
‘As AI rapidly commoditizes code generation, the strategic leverage shifts upstream—to product teams and the decisions they make. Thinkpilot is uniquely positioned at this inflection point, giving product managers a 10x edge through a purpose-built AI co-pilot that integrates directly with their workflow. With specialized agents for PMs, Thinkpilot transforms scattered thinking into structured output,’ LAUNCHub Ventures’ founding partner Todor Breshkov comments.
Narrow Focus with Broad Collaboration Potential and Scalable Human-First Onboarding
Speaking of possible scaling of Thinkpilot’s solution outside product development, Mr Veselinov underlines that the team won’t dilute its focus before it nails the application in product development to the point of perfection. This means that from workflows to data models, everything is optimized for product management. That said, Thinkpilot naturally does support collaboration with teams like marketing or CS, especially when they’re involved in product initiatives, framing those use cases through the lens of product development.
A smooth onboarding process for users who may not be familiar with AI-driven tools has been a major focus for the Thinkpilot team. Sitting in a new category itself, the platform guides users through onboarding with a human-first approach. Early on, the team is hands-on, chatting directly with users, answering questions, and watching how they work. Scalable onboarding is also in place: short videos, in-app nudges, and a co-pilot that acts like a real teammate.
Megan Wolff, Co-Founder and Head of Growth at Thinkpilot
‘You don’t have to ‘learn’ Thinkpilot, it proactively suggests what to do next, creates tasks, and interacts like another member of your team. That familiarity makes it feel less like a tool and more like a collaborator,’ Ms Wolff reassures.
Thinkpilot already received critical early feedback. According to Mr Veselinov, the first prototype proved the concept’s potential, but the second one revealed that the used technology would hinder scaling. The startup had to make a major pivot in its tech stack and rethink designing the product experience. What remained is the product vision, to empower product teams to turn ideas into value-driven solutions swiftly.
Critical AI Support for Modern Product Teams
‘The AI workspace category is growing fast, and we believe Thinkpilot is poised to lead in one of its most valuable verticals. We’re excited to support this team as they redefine the future of product development,’ Mr Breshkov summarizes.
The majority of the newly raised investment fuels Thinkpilot’s senior, experienced team that can finalize the product that the founding team envisioned and get it to product-market fit. Growth is another key priority: talking to the market, running feedback loops, and getting Thinkpilot into the hands of real-life teams.
‘Everything we’re doing, whether it’s product, design, or outreach, is focused on learning fast, refining fast, and reaching the teams who need us most,’ Mr Veselinov wraps up.
As AI continues to reshape how products are built, its largely untapped value lies in enhancing the strategic decisions that shape products. For product teams, AI assistance offers a way to cut through the unnecessary complexity, align stakeholders, and move from ideas to outcomes with greater clarity. Tools like Thinkpilot highlight how purpose-built AI can become an integral partner in this process.
Kostiantyn is a freelance writer from Crimea but based in Lviv. He loves writing about IT and high tech because those topics are always upbeat and he’s an inherent optimist!
