Anari AI develops a software stack that delivers cloud-based virtual chips and software-defined accelerators to optimize AI/ML workloads, including FPGA customization for domain-specific performance. The technology aims to reduce compute requirements and associated carbon emissions by enabling efficient use of CPUs, GPUs, standardized accelerators and ad-hoc customized hardware. The team highlights use cases such as processing massive 3D point clouds for digital twins and infrastructure planning and plans to focus on a few initial verticals for early commercial traction. Anari AI intends to use its latest funding to further develop the product and continue exploring suitable verticals. The company was founded in 2020 and is based in Serbia. The founding team includes CEO Jovan Stojanovic, who has extensive AI/ML project experience, and the startup is supported by a group of advisors and business angels. Anari AI offers a cloud platform that lets companies design personalized AI chips with one click, shrinking hardware development from 12–18 months to weeks. The platform is built on a Python-based programming framework developed by Anari that opens hardware design to software engineers. Its first configurable cloud chip, Thor X, delivers 100x more efficient processing of 3D point cloud and graph data structures compared to a high-end GPU. The company says it will use the new funding to continue developing the platform and bring it to market. Anari's team includes Ph.D. researchers, machine learning experts, and senior software engineers led by co-founders Jovan Stojanovic (CEO), Stefan Sredojevic (COO), and Bogdan Vukobratovic (Head of R&D). The company is based in Novi Sad, Serbia, and maintains an office in San Francisco; advisors and board additions include industry veterans such as Roland Manger and Sasha Ostojic.

Total raised

$5M

Funding rounds

2

Latest round

Seed

Latest activity

Jan 2023