Belfort provides purpose‑built hardware and software that accelerates Fully Homomorphic Encryption, enabling real‑time computation on encrypted data without ever decrypting it. Its accelerator chips and optimized FHE stack deliver sub‑second latencies at scale, available as a cloud service on AWS or on‑premises appliances for enterprises in finance, government, healthcare, and crypto.
Funding
$7.9M raised to dateRaised to date based on public sources. This may differ from the amount the company actually raised and is based only on what is publicly available on the internet.
ACNJDNR+2Founders
Product
Problem
Organizations handling sensitive data must decrypt information before processing it, exposing the data to potential breaches, insider threats, and regulatory violations. Existing secure execution methods, such as trusted execution environments, still rely on trusted hardware and can be vulnerable to hardware exploits.
Solution
Belfort provides purpose-built hardware and software that accelerates Fully Homomorphic Encryption (FHE) to enable real-time computation on encrypted data without ever decrypting it. By integrating custom acceleration chips with optimized FHE algorithms, Belfort reduces processing times from minutes to sub‑second latencies, making encrypted compute practical for production workloads. The platform is available both as a cloud service on AWS and as on‑premises appliances, allowing seamless integration into existing pipelines while preserving end‑to‑end data confidentiality. Belfort’s solution abstracts the complexity of FHE, offering developers a drop‑in accelerator that works with standard FHE libraries and delivers scalable performance for high‑stakes applications.
Target Audience
Primary customers are enterprises in finance, government, healthcare, and crypto that need to process sensitive data—such as transaction records, patient/genomic information, or confidential intelligence—while maintaining strict privacy and compliance requirements.
Features
- Dedicated accelerator hardware (Velox FPGA/ASIC) designed specifically for TFHE operations, delivering orders‑of‑magnitude speedups over CPU implementations
- Optimized FHE algorithmic stack, including patented streaming architecture and algorithmic refinements that lower latency for workloads such as fuzzy name matching
- Cloud deployment on AWS for instant, on‑demand scaling without managing physical hardware
- On‑premises appliance option for strict data‑sovereignty, regulatory compliance, and dedicated performance guarantees
- Compatibility with existing FHE libraries (e.g., TFHE‑rs) requiring minimal code changes to leverage acceleration
- End‑to‑end encryption guarantees: data remains encrypted throughout computation, eliminating reliance on trusted hardware enclaves