About the company


Members type

1015 Lausanne, Switzerland

Company size

Aldo Podesta



  • Clean water and sanitation

  • Affordable and clean energy

  • Industry, innovation and infrastructure

  • Sustainable cities and communities

  • Responsible consumption and production



Alliance member


L2F aims to bring mathematical innovation to machine learning and to deploy best quality ML software in corporations.

L2F is a Swiss technology company based on the EPFL innovation park operating since 2017. Heavily active in mathematical research, the goal of L2F is to improve the quality of state-of-the-art Machine Learning and Optimisation tools by incorporating concepts from algebraic topology and geometry (topological learning). In partnership with EPFL's lab for topology and neuroscience, recipient of a 550K Innosuisse grant, L2F makes a selection of its Topological Learning functions available open source through a Python library (Giotto). Commercially, L2F's business model is an "open core" one, as it licenses commercial software modules to corporate clients. L2F's commercial modules were built using its core technology (Giotto) as well as its B2B client data and expertise. L2F currently licenses 8 different software modules to bring impactful solutions to corporate, societal and environmental problems. Modules include image segmentation, churn analysis, demand forecasting, logistical optimisation, B2B market assessment...etc. To day, L2F licenses a dozen of modules to world famous corporations. One of the long term ambitions of L2F is to specialise in high impact solutions that have a positive impact on our environment. Two of our current projects qualify in that regard in the maritime shipping and agri-tech industries. L2F counts 30 employees and has raised CHF 3M in Series A funding and 650K in governmental grants.



Optimize transportation routes, energy consumptions, adjust quantities to needs, make accurate forecasting

Available in

Categories of application
Clean water and sanitation, Affordable and clean energy, Industry, innovation and infrastructure, Sustainable cities and communities, Responsible consumption and production

Sector of application
Water distribution and use, Energy distribution & management, Industrial processes, Freight transportation, Urban and inter-urban mobility, Agriculture and farming for food production, Primary resources management


Bakis logistical planning engine

Offer Bakis logistical planning engine

A centralised system which makes live container repositioning decisions to make transport networks cleaner and efficient

Bakis uses advanced techniques of time-series forecasting to predict the important variables of shipping networks, in particular by integrating external data in a user friendly way. The predictive outputs (such as future demand for containers across the network) are fed into a flexible and powerful optimisation engine with the objective of generating optimal container allocation, route and traffic decisions. Such decision model is highly customisable and can maximise different variables such as profit but also energy consumption or CO2 emissions giving shipping company the possibility of controlling such metrics. The Bakis solution differentiates to existing logistic optimizers thanks to its unique predictive component and its efficient computational architecture. The economic value added of Bakis counts in hundreds of millions if fully implemented in one of the top 5 shipping companies.

Available in

Categories of application
Industry, innovation and infrastructure

Sector of application
Freight transportation