Neural DSP

Neural-network amplifier modelling: the Quad Cortex floor unit (quad-core SHARC DSP) and plugin suites run controllable full circuit models, and Neural Capture learns and replicates any physical amp in real time. The company publishes its own research on posing controllable amp modelling as a machine-learning problem — a data-driven pipeline that matches circuit-simulation quality without circuit analysis. TINA (Telemetric Inductive Nodal Actuator) combines robotics with ML to digitise analog gear with no human in the loop. Founded 2017 in Helsinki by Douglas Castro and Francisco Cresp. End market is musical instruments, but the core is real-time ML on custom DSP hardware.

The row

Sector

AI & Data

Stage

Scaled — International revenue over EUR 20M and not declining

Location

Helsinki

Founded

2017

Business ID

2847533-1 · Finnish trade register

Funding

~$6.7M (2020, CapMan, Trind, Rubylight); ~114 employees

Links

Where this sits on the map

Neural DSP is one of 11 companies in AI & Data. The others:

How this row was decided

Every row on this map is checked by hand against the trade register and the company's own material, and placed in the sector that best describes its business or its technology. Service firms and consultancies are left out, however advanced the technology they work with — the test is where the revenue comes from. Only funding figures the company or its investor has published are shown. The full rules are on the landscape ↗