M&A and Fundraising Analysis for October 28, 2026

AQQI: real estate purchase as an organizational problem, not a search problem

Source: jaimelesstartups.fr → · Sector Construction & PropTech — 📬 subscribe to the Construction & PropTech newsletter

AQQI is a French startup founded in early 2026 by Alexandre Ducoeur Martins and Julie de la Porte des Vaux, both former employees of Studapart — a real estate rental platform sold for over 55 million euros. AQQI is developing a personal AI assistant dedicated to real estate purchases: the tool is designed to understand the buyer's project, track its evolution over time, and handle the administrative and relational steps of the process — follow-ups, visit organization, information gathering, coordination with professionals. According to the company, buyers spend an average of 150 hours on their project and exchange over a hundred times by email and phone; AQQI claims to reduce the time-consuming tasks of this process by 75%.

The startup announced a 4 million euro pre-seed round, led by Axeleo and Aquiti, with the participation of Kima Ventures, Pierre Kosciusko-Morizet, and 50 Partners. The solution is being tested with several dozen users in Île-de-France, Bordeaux, and Marseille; deployment is planned for 1,000 buyers before a public launch in early 2027.

What distinguishes AQQI's positioning from listing portals or property search tools is the entry point: AQQI does not seek to help the buyer find the right property — it seeks to prevent them from getting bogged down in the process once they have decided to search. It is a project management assistant, not a search engine. The addressable market is large — every buyer in France — but monetization and retention remain to be proven: a real estate purchase is a rare event in a lifetime, which requires either a transaction-based model or an extension to other stages of the journey (financing, renovations, resale).

The founders' credibility in the real estate market is real, and the pre-seed ticket is consistent with the stage. The challenge in the coming months will be less about convincing users than about demonstrating a conversion rate and value high enough to justify the acquisition cost in a market where buyers do not behave like subscribers.

📩 Subscribe to our newsletter to follow daily M&A and fundraising news: https://proplace.co/newsletter

Atlas Bio: simulating clinical trials before launching them

Source: tech-insider.org → · Sector HealthTech & Digital Health — 📬 subscribe to the HealthTech & Digital Health newsletter

Atlas Bio is a Basel-based startup building an AI model capable of simulating clinical trials — what researchers call in silico trials — to predict, even before recruiting a single patient, how a drug candidate will behave in Phase II, at the individual level rather than a population average. The problem it tackles is quantifiable and brutal: approximately 70% of drugs fail in Phase II, the stage where human trials truly begin, after years and hundreds of millions spent in preclinical development.

The company announced 12.5 million dollars in total funding, comprising a 10 million dollar seed round led by Creandum — whose partner Katharina Wilhelm joins the board of directors — and a previously raised 2.5 million dollar pre-seed round led by First Momentum Ventures and LEA Partners, revealed simultaneously. The funds will be used to strengthen scientific and commercial teams, deepen platform validation, and expand the corporate client base by the end of 2026 — Atlas Bio already describes itself as selling to key accounts.

The surface reading: a deeptech biotech startup raises a modest amount in a week dominated by AI infrastructure checks of a completely different caliber.

The twist lies in the nature of the problem. Phase II clinical trials are not just costly — they are the primary destroyer of value for the entire global pharmaceutical industry. Each failure at this stage represents years of R&D and hundreds of millions swallowed. What Atlas Bio offers is not a productivity tool: it's a filter placed before the abyss. If the platform can predict with sufficient reliability which candidates will fail — and which deserve to advance — it is not selling software; it is selling insurance on research capital. The business model targets pharmaceutical laboratories and biotechs investing in molecule pipelines: for them, paying for a simulated trial is rational as long as it reduces the probability of burning ten times more in a real trial.

The real validation challenge is technical: an AI model for clinical trials is only as good as its training data and its ability to generalize to new molecules. The question that key account clients will ask is not "does your AI predict well?" but "what data did you train on, and how do you prove it holds for my specific molecule?"

For an industrial partner or a corporate pharmaceutical investor, Atlas Bio represents exactly the type of asymmetric bet that the ticket size makes accessible: if the validation holds, the value created is immeasurable compared to the 12.5 million committed.