Cosmo Tech: our review
Cosmo Tech builds digital twins of supply chains or asset fleets to simulate the effect of a decision before making it: opening a site, changing a supplier, sizing an investment. The vendor, born in Lyon in 2010, runs a cloud platform on Microsoft Azure.
What we take away from it
What counts in its favor
- The simulation engine (CoSML) models an entire system — flows, stocks, constraints — rather than a single isolated indicator, letting you test a scenario before running it for real.
- About twenty major accounts cited by the vendor (RTE, Alstom, Nexans, Renault, SNCF), indicating real production use rather than just pilot projects.
- An information security policy aligned with the ISO 27001 standard, publicly documented — rare for this kind of B2B vendor with a long sales cycle.
What counts against it
- The platform is presented as cloud-based, deployed on Microsoft Azure: no on-premises option on the client's own servers appears in the public documentation.
- No pricing page: the model is a quote after a demo, with no order of magnitude published anywhere.
- Ownership changed in late 2024 (a US fund came in, new CEO appointed): worth confirming with the vendor whether this changed the product roadmap or the hosting strategy.
A serious tool for a large organization that wants to price out a logistics or industrial decision before committing to it, provided it accepts cloud processing and a case-by-case negotiated rate.
The facts
- Data hosting
- Hosting not verified
- Categories
- Manufacturing,Data analysis
- Data verified on
- 30 July 2026
What happens to your data
- Is your data used to train their models?
- Subprocessors
- Not named individually. The privacy policy mentions transfers to subprocessors located outside the European Union, including the United States, and cites Microsoft (Azure) for audience measurement.
- Retention
- For a client under contract: the contract term plus three years. Recruitment files are kept two years, accounting data ten years.
Cosmo Tech in detail
Cosmo Tech sells simulation software that builds a digital twin of an entire system — a supply chain, a fleet of industrial assets, a distribution network — to test the effect of a decision before executing it for real. The vendor was born in Lyon in 2010, spun off from ENS de Lyon, and claims about twenty major accounts among its clients.
What it actually does
A client models its system with CoSML, a proprietary language for modeling complex systems: flows, stocks, capacity constraints, lead times. The platform then runs thousands of scenarios — a supplier closing, demand doubling, a site opening elsewhere — and compares the results. RTE, Alstom, Nexans, Renault, and SNCF are among the references cited by the vendor, which places the tool on the side of the large industrial or energy account rather than the SMB.
Why it interests us
It's a good example for settling a question that keeps coming up in industry: is the computation done at the client's site or the vendor's? Here, the answer is clear and documented: the platform is presented as deployed on Microsoft Azure, with no on-premises option on the client's own servers visible in the public documentation. For a digital twin built from flow and capacity data — potentially sensitive for an energy company or a defense industrial firm — that's a factor to weigh before committing. See our AI for industry page. On the actual share of artificial intelligence involved: the core of the product remains a simulation and optimization engine, a discipline that existed before generative AI; machine-learning components are added on top without replacing the modeling engine.
Where it falls short
Three points to watch. The first is hosting: everything runs through the Azure cloud, with no documented local-hosting option — worth confirming directly with the vendor if that's an exclusion criterion for your organization. The second is price: no pricing page, a quote after a demo, with no order of magnitude communicated in advance. The third is ownership stability: in late 2024, a US fund entered the capital and a new CEO was appointed, a change whose effect on the product roadmap isn't documented publicly as of this fact sheet's verification.
Our recommendation
A sound choice for a large group that must decide on a heavy investment and that accepts cloud processing of its flow data. Avoid it if local hosting is a prerequisite, or if you're looking for a posted price to quickly compare several vendors.
Who it's for, and who it isn't
Suited for
- Large industrial or energy group that must decide on a heavy investment (new line, new site, supplier change) and wants to price several scenarios before deciding
- Supply chain leadership that wants to simulate the effect of a supply disruption before it happens
Not a fit
- SMB looking for a self-service tool at a posted price: here every project goes through a demo, then a quote
- Organization that requires simulation data to stay on its own hardware, with no pass through a public cloud
Frequently asked questions
- Does Cosmo Tech run on our own servers or in the cloud?
- In the cloud. The vendor presents its platform as deployed on Microsoft Azure, with no on-premises option documented publicly. Simulating a supply chain or an asset fleet therefore means a data flow to that cloud.
- How much does a Cosmo Tech simulation cost?
- No pricing is published. The vendor works by demo then quote, a common model for this kind of custom project sold to large industrial accounts.
- Does Cosmo Tech really use AI, or is this classic simulation?
- Both, mixed together. The core of the product is a simulation and optimization engine based on complex-system modeling (CoSML), a discipline older than generative AI. Machine-learning components are added to train models and enrich monitoring, but the exact split between the two isn't detailed publicly.
- Is the company stable?
- It has existed since 2010 and claims about thirty million euros raised in total, but its ownership changed in late 2024 with a US fund coming in and a new CEO appointed — a point worth clarifying before a long-term commitment.
- What's the difference with Sight Machine or Braincube?
- Sight Machine and Braincube start from real production data to analyze what's already happening on a line. Cosmo Tech starts from a model of the system to simulate what would happen if a decision were made — before making it.
Alternatives to Cosmo Tech
Manufacturing · Data analysis
Sight Machine
Digital twin of production, to spot where output is being lost.
Manufacturing · Data analysis
Braincube
Industrial process optimization from machine data.
Manufacturing · Data analysis
Cognite
Industrial platform that structures plant data (sensors, ERP, historians) to deploy AI applications.
Our sources
- Platform pageaccessed on 30 July 2026
- Privacy policyaccessed on 30 July 2026
- ISMS Management page (information security)accessed on 30 July 2026
- About pageaccessed on 30 July 2026
Prices and hosting terms change often. We recheck them periodically, but the publisher's page remains the only up-to-date reference.