Fast, research-backed, and built to scale. Our advantage lies in our agility, competitive pricing, and readiness to get the job done. Closely linked to current tech research, we provide state-of-the-art solutions backed by solid project experience. Our network throughout ETH and UZH and diverse technical interests allow us to cover broad IT domains. We support companies in custom AI automations (e.g., via LangGraph AI Agents), Proof-of-Concepts, machine learning, and (Cloud) infrastructure.
BSc. Computer Science ETH (enrolled in MSc.)
Machine Intelligence and Data Management Systems
BSc. Computer Science UZH (enrolled in MSc.)
Software Systems / Banking and Finance
Custom model-agnostic and open-source AI agent (LangGraph) for validating and enriching company data.
Outcome: Fully automated the retrieval of live company data in a data sovereign manner using self-hosted models. More info: GitHub
A server/client to test multiple compute piplelines including large scale data transfer. We compared: h1/h2 in large data transfer, for computation WebAssembly, WebGPU and plain JavaScript.
Key findings: For the frontend WASM performed the best and for speed optimization it is best to compute on the server side.
Extention of a 3D Gaussian Splatting codebase to allow for training on deformable meshes.
Outcome: This enabled photo-realistic visualization of expected results of surgical procedures for teeth alignment in real time.
Core idea: Can we connect IPC streams or even RPC technology with the browser? Could it make data transfer faster or more efficient? The implementation consisted of an Apache Arrow Flight server that sends IPC streams on the gRPC format.
Key findings: Because of browser side limitations the performance was not increased.
fn introduces a function that translates challenges into results. When we founded fn-labs, we wanted to bring precisely this way of thinking to the world of consultancy. Companies today face technological challenges: whether it’s the productive use of artificial intelligence or the development of modern data pipelines. We translate this complexity into clear, functional solutions. We develop MVPs and feasibility studies to bring your IT and AI visions to life in a precise and agile manner.
Proposal
Contact
Tell us what you want to improve, automate, validate, or build. We will respond with a concrete next step and a first assessment call.