M&A and Fundraising Analysis for October 13, 2026

Pinson acquires Maison Verdier: a family SME empowering itself for the long run

Source: agence-api.ouest-france.fr → · Sector Industrial Tech & Manufacturing — 📬 subscribe to the Industrial Tech & Manufacturing newsletter

La Confiserie Pinson is a Vendée-based family business founded in Nantes in 1895, located in Montaigu-Vendée since 2011, with 30 employees and €3.75M in turnover. It manufactures confectionery. Maison Verdier is a Gironde-based chocolate factory created in 1945, with 7 employees and €1.5M in turnover, known for its "raisins dorés" — Sauternes-soaked raisins coated in 70% dark chocolate. Pinson acquired Maison Verdier on October 1; the historical recipe will be maintained. The group also announced its intention to double the size of its factory within two years.

A modest-sized operation, but with sound logic: Pinson expands its range into specialty chocolate without starting from scratch, inherits a brand with a strong identity and artisanal know-how, and gains the industrial capacity to absorb both activities. The simultaneously announced factory doubling signals that the acquisition is not defensive — it's a bet on growth.

Koddex raises €3M to unite industrial engineers on a single platform

Source: lagazettefrance.fr →

Koddex is a French startup founded in 2024 by Stéphane Dicostanzo, a former software engineer at Exotec. It develops what it describes as an operating system for teams designing complex industrial products subject to high regulatory requirements: aerospace, space, defense, medical, robotics, energy. Its platform models the complete architecture of a product — components, attributes, relationships, requirements, design rules — and brings together data, documents, tests, versions, decisions, and maintenance operations in a common environment. A knowledge graph allows for cascading identification of the consequences of a modification on the entire system. The platform also integrates regulatory repositories (ISO 9001, IEC 61508, AS9100, ISO 13485, ISO 26262) to document decisions from the design stage. It is also preparing for the arrival of AI agents via the Model Context Protocol, with logging of each action.

Koddex raises €3M from Breega (lead) with the participation of Exotec, a warehouse robotics specialist — from which the founder originated. The funds will finance team strengthening, enrichment of software integrations, and commercial development in France and Europe. The startup has joined EuraTechnologies' The Square program for six months.

Exotec's presence in the capital is not insignificant: it is both an industrial validation by a potential typical client and a gateway to a network of integrators and equipment manufacturers. In the sectors targeted by Koddex, the problem of dispersed engineering tools is real and costly — the value of a platform that centralizes without replacing each specialized tool is all the greater as regulatory requirements increase.

Finches raises €2M to alert agricultural buyers before the harvest is compromised

Source: eu-startups.com → · Sector Horizontal & Productivity SaaS — 📬 subscribe to the Horizontal & Productivity SaaS newsletter

Finches is a Bavarian startup founded in 2025 by Catharina van Delden (CEO) and Dr. Stefanie Glenn (CTO), joined in 2026 by Alexandra Vázquez Bea (CFO/COO). It develops an early warning platform for agricultural raw material procurement teams in the agri-food, pharmaceutical, and cosmetics sectors. Its platform aggregates field data, internal supply data (supplier profiles, cultivated areas, contracts), and external signals (weather, satellite images, local news, field agronomist reports) to warn buyers of supply chain disruptions several weeks before they materialize, with concrete recommendations: activate alternative suppliers, buy earlier, reschedule production.

Finches raises €2M in pre-seed funding, a round led by High-Tech Gründerfonds (HTGF) and co-led by Vanagon Ventures, with participation from Bayern Kapital and the incubator UnternehmerTUM Funding for Innovators, as well as a syndicate of industrial and tech angels.

The timing is telling: the fundraising closes at the end of a European summer marked by drought. Finches does not sell a retrospective reporting tool — it sells reaction time to buyers whose traditional systems only detect disruptions when it's too late to act. In sectors where climate volatility is now structural, the value of an actionable early warning signal is directly measured in avoided production shutdown costs.

Haven: $15M to build a personal digital twin that acts on your behalf

Source: dealroom.co →

Haven is an American startup developing what it calls a "personal digital twin": an AI assistant that learns a user's habits, preferences, and decisions through their interactions — calls, emails, SMS, apps — and proactively acts on their behalf, without waiting to be prompted. Unlike classic assistants that forget from one session to the next, Haven builds a persistent memory graph for each user. In practice: preparing daily briefings, managing emails, reminders, phone calls, processing e-commerce returns. Since its private beta, the platform has recorded over two million interactions.

The startup, founded by Frank Addante and Duc Chau, emerges from stealth with $15M in pre-seed funding, a round led by Mayfield.

The funds will primarily be used to build a developer platform allowing third parties to connect their applications to the user's digital twin, with a planned extension to physical devices (televisions, cars, connected objects).

Mayfield's thesis is explicit: the next wave of consumer AI will not be chatbots but agents that know the user and act for them. Haven bets on the portability of personal context — the interface changes, the memory remains. It's a bet on the identity layer of personal AI, not on a particular product. If the developer platform works, Haven could become the infrastructure on which other applications connect — a powerful distribution model, but one that requires solving issues of trust and privacy that $15M alone will not be enough to address.

Songbot raises $14.9M to industrialize robot tennis training

Source: dealroom.co → · Sector Education & EdTech — 📬 subscribe to the Education & EdTech newsletter

Songbot is a Chinese intelligent sports equipment company. It has developed autonomous service robots powered by its own AI model specialized in sports, covering visual perception, motion analysis, and training decisions. Entering the market with table tennis, it has adapted its technology to tennis, and then aims for other sports. Its PACE robot raised over $2.7M on Kickstarter in October 2024; its all-in-one Aura robot exceeded $1M in crowdfunding in five hours at its launch in May 2026 and ultimately reached over $4M. Songbot claims over 300,000 users worldwide and two billion recorded services.

It is raising $14.9M (approximately $13.8M USD) in a Series B round, led by the Shanghai State Investment Pilot Fund, with participation from Yuanhe Puhua, Pudong Venture Capital, and long-standing shareholder BlueRun Ventures. The funds will finance R&D on the AI sports model, motion vision, training decisions, international expansion, and new products for other sports. On September 28, 2026, Songbot became an official service robot partner of the Australian Open and a supplier to the Australian National Tennis Academy.

The coach-to-player ratio cited (800 active players per coach in the United States, according to the International Tennis Federation) illuminates the thesis: Songbot does not compete with elite professional coaches; it targets the vast majority of players who do not have access to them. The partnership with the Australian Open is less a commercial contract than a validation of international credibility — a useful signal for expansion outside China.

I2Cool: $14M to cool data centers in the AI era

Source: af.net →

I2Cool is a startup specializing in AI-powered cooling solutions for data centers. It announced at the Ai Everything summit in Abu Dhabi that it has raised $14M, with the ambition of reducing data center energy consumption by up to 40%. The fundraising attracted strategic investors who share the vision of a more energy-efficient computing infrastructure, in a context where the United Arab Emirates seeks to establish itself as a regional AI hub.

The provided context does not detail the investors or the precise cooling technology. What is established: the demand for this type of solution is structural — the computing loads related to AI consume quantities of energy that make cooling a critical item, both for operational costs and for the environmental commitments of data center operators.

In a sector where almost all the money goes to chips and models, a company that tackles the energy efficiency of the infrastructure itself occupies a discreet but potentially sustainable position: its clients are not AI startups but cloud operators and sovereigns building computing capabilities — a much more stable clientele.

Doris Labs raises $600K to make commercial data readable by AI agents

Source: startup.eu →

Doris Labs is a startup based in Felixstowe (UK) that develops a semantic layer for sales teams: it indexes sales calls, emails, CRM data, internal documents, and prospecting activities into structured objects that AI systems can interpret. The goal is to help sales teams understand how opportunities are progressing and to automate parts of the sales process. It notably develops AI agents for sales automations and a "Sidekick" tool designed to work on each individual deal.

Doris Labs raises $600K (approximately €535K) in pre-seed funding, a round led by Fuel Ventures, with participation from Manchester Angels and other business angels.

A modest startup operation in a segment — AI applied to sales operations — that attracts many players. Doris Labs' differentiation relies on the upstream semantic layer, which structures data before entrusting it to agents, rather than having AI act directly on raw and fragmented data.

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