FIXER launches Sovereign GaiXer, a Lenovo-co-developed on-premise generative AI workstation for healthcare and manufacturing.
The AMW Read
FIXER extends its existing GaiXer platform into a Lenovo-co-developed on-prem hardware appliance for regulated verticals, an incremental product-line update rather than a new capability or entrant.
FIXER launches Sovereign GaiXer, a Lenovo-co-developed on-premise generative AI workstation for healthcare and manufacturing.
FIXER, developer of the GaiXer generative AI platform, released Sovereign GaiXer, a palm-sized on-premise workstation built with Lenovo that delivers roughly 1 petaFLOPS of compute, described as equivalent to about 1,000 laptops. The appliance connects directly to a company's existing shared folders to read PDF, Word and Excel files without extra data preparation, answers questions with citations back to the source documents, and auto-drafts business documents such as quotes, reports and meeting minutes. All processing runs inside the customer's own hardware, so no input data leaves the premises. Pricing is a one-time payment of 3,806,400 yen (~$25,700, tax excluded) covering the GPU hardware, software license, and a three-year warranty, with no recurring subscription fee; FIXER says deployment can be same-day.
The product targets healthcare, manufacturing, finance and government buyers, the exact segment that has been slowest to adopt generative AI because of data-security policy and unclear use cases. Framing the appliance as "sovereign" — keeping data inside Japan and inside the customer's own walls — turns on-prem deployment into a packaged, buyable SKU rather than a custom integration project, competing against cloud-subscription incumbents on a capex-versus-opex basis.
FIXER cites its existing GaiXer Medical Agent, which drafts discharge summaries from electronic health records, reporting roughly 1,000 staff-hours saved over three months at one hospital, with about 90% of nearly 200 surveyed physicians reporting efficiency gains and 80% reporting satisfaction; that agent is now deployable on the sovereign hardware. For vendors selling into regulated verticals, bundling dedicated inference hardware with a software agent is a distribution path worth watching alongside pure cloud-API offerings.