Skip to main content
Back to News
Desilo secures research support for encrypted LLM inference
Technology
2 min read
KR

Desilo secures research support for encrypted LLM inference

The AMW Read

The award advances Desilo's existing encrypted-inference research, but the report provides no performance or deployment evidence to establish broader infrastructure impact.
NoveltySignificance
AI Infra · Player Map

Desilo secures research support for encrypted LLM inference

Desilo said on October 28 that it signed a Deep Tech TIPS agreement on October 23. The program provides up to 1.5 billion won in government research funding over three years. Recommended by BonAngels Venture Partners, Desilo will develop a homomorphic-encryption LLM inference engine and confidential-data AI services. This follows its selection for the regular TIPS program in 2020.

The project targets the infrastructure layer where models process sensitive inputs. Fully homomorphic encryption allows computation on encrypted data without first decrypting it. Desilo aims to let users access AI services without revealing plaintext to the computing party. It plans to advance its GL Scheme encryption technology and THOR encrypted LLM inference technology. According to the company, two GL Scheme papers were accepted at CRYPTO 2026. THOR, jointly developed with Professor Miran Kim's team at Hanyang University, was presented at ACM CCS 2025. These research milestones support the technical basis; the funding announcement does not establish commercial readiness.

For builders and investors, the concrete question is whether encrypted inference can meet deployment requirements in finance, healthcare, and public services, where Desilo plans to expand applications. Evaluation should focus on inference latency, operating cost, and model compatibility. The report provides no measurements for those criteria or customer deployment results. The award supports further development, while the engine's practical value remains to be demonstrated.

#Desilo #AIInfrastructure #HomomorphicEncryption #LLMInference #DeepTechTIPS

#Desilo#fully homomorphic encryption#LLM inference#Deep Tech TIPS

How This Connects

Based on AI Infra · Player Map

  1. 3w agoNAVER sets November 6 payment date for NVIDIA's $1 billion AI factory investmentNAVER
  2. 3w agoDesilo secures research support for encrypted LLM inference · THIS ARTICLE
  3. 3w agoNVIDIA opens Vera Rubin rack designs for denser AI data centersNVIDIA
  4. 1mo agoNscale's $4.19B H1 financing combines Series C equity and data-center debtNscale

More news from DESILO

Stay updated with the latest news and announcements from DESILO.

View all DESILO news

Discover AI Startups

Explore 5,000+ AI companies with VC-grade analysis, funding data, and investment insights.

Explore Dashboard