Reduce energy consumption across storage, networking, and processing infrastructure.
Innovative · Efficient · Secure
Cut storage and transfer costs to the extreme with zero loss
Our patented technology reduces files already compressed with ZIP, LZ4, Zstd, and others. You store and send fewer bytes, so costs go down, transfers run faster, and energy use drops.
Immediate impact, drop-in
Breakthrough technology that redefines what's possible in data compression
Dramatically reduce your infrastructure expenses with unprecedented compression ratios.
Works on ZIP, LZ4, Zstd, and other pre-compressed formats where traditional methods fail.
Perfect reconstruction guaranteed every single bit recovered exactly as original.
Explore the key moments
-
2023
- Apr: Granted Patent
- Aug: Founded
- Nvidia Inception
-
2024
-
2025
A short demonstration of how the compressor works in practice:
Today’s data problem
- Scale: petabytes across clusters/regions; backup windows slip and restores lag.
- Budget: capacity growth and cross-region egress push costs up on-prem and in cloud.
- Energy: every byte burns power; efficiency targets keep tightening.
Why existing approaches fall short
- Classical lossless codecs (ZIP/LZ4/Zstd) hit limits on high-entropy logs, binaries, and floats.
- Lossy methods change the payload — unacceptable for backup, audit, and science.
- Dedup/tiering reorganize data but rarely shrink the bytes that must actually move.
Revolutionary Use Cases
Transform industries with unprecedented compression capabilities
Data Centers
Optimize storage density and reduce infrastructure costs
Transform data center economics with up to 99.99% compression. Reduce rack space, cooling costs, and energy consumption while maintaining lightning-fast access speeds.
Integrates seamlessly with existing storage stacks and file systems.
Unlock higher storage density without re-architecting applications.
Key Benefits:
- Reduced infrastructure costs
- Lower energy consumption
- Improved storage density
Backup & DR
Accelerate backup processes and reduce storage requirements
Accelerate backup windows and shrink retention footprints without sacrificing fidelity.
Slash replication bandwidth and reduce backup windows even for large datasets.
Maintain bit-exact integrity for audit, compliance, and rapid verification.
Key Benefits:
- Faster backups
- Shorter restore times
- Lower storage footprint
Transfer & Edge → Cloud
Minimize bandwidth usage and transfer times
Optimize edge-to-cloud transfers with unprecedented reduction to lower bandwidth costs and improve application performance.
Cut egress charges and time-to-first-byte for bulk migrations and syncs.
Works even on pre-compressed archives and logs across heterogeneous links.
Key Benefits:
- Reduced bandwidth costs
- Faster transfers
- Improved edge performance
Video Games
Reduce game asset sizes and download times
Transform distribution with smaller downloads and patches; compact logs and multiplayer telemetry.
Speed up first-time installs and day-one patches across platforms and storefronts.
Reduce CDN egress while maintaining asset fidelity and player experience.
Key Benefits:
- Smaller downloads
- Faster updates
- Reduced distribution costs
AI Models & Datasets
Compress training data and model distributions
Compress training data and model distributions to speed up experimentation and reduce storage.
Shorten training cycles by alleviating I/O bottlenecks on checkpoints and tensors.
Bit-exact compression preserves numerical fidelity for reproducible research.
Key Benefits:
- Lower storage
- Faster IO
- Scales with dataset size
Satellite Communications
Maximize throughput in bandwidth-constrained environments
Maximize throughput in constrained environments, sending more data per link budget.
Increase effective throughput on high-latency, low-bandwidth satellite links.
Deliver more telemetry and imagery within the same pass window.
Key Benefits:
- More data per window
- Lower OPEX
- Better real-time visibility
Performance snapshots
Real-world–inspired examples; actual results vary by data type, structure, and pipeline.
Dataset | Input Size | Output Size | Throughput (input GB/s per node) | Notes / Hardware |
---|---|---|---|---|
Telemetry logs (already Zstd) | 12.0 TB | 360 MB | 1.6 GB/s | Dual 32-core x86, NVMe; inline agent |
Binary artifacts (mixed) | 3.7 TB | 111 MB | 1.2 GB/s | x86 cluster; batch |
Object payloads (various) | 820 GB | 25 MB | 0.9 GB/s | ARM edge nodes → cloud |
- Lossless pipeline; no changes required to your file/object/stream interfaces.
- Benchmarked on representative workloads (research clusters and enterprise clients) under NDA.
- Validation: SHA-256 before/after, randomized spot checks, and fault-injection on inline paths.
- Environments: dual-socket x86 (64 cores), 256 GB RAM, NVMe; and ARM edge nodes.
- Baselines: Zstd/LZ4 and ZIP where applicable.
Solutions & delivery options
Production-ready across on-prem, cloud, and air-gapped environments. Security: processing runs inside trusted execution environments (TEEs). Licensing: a valid license is required for all products and deployments.
SPIRAL-BOX
Portable, offline hardware for secure, lossless compression
- No internal storage, operates solely via external drives or connections
- Strong encryption; no internet connectivity
- Cuts storage footprint and transfer overhead without quality loss
Managed Enclave Service
Cloud-hosted, fully managed compression service
- Zero infrastructure management
- Global availability
- Pay-per-use pricing
- Requires internet connectivity; not available offline
FPGA Offload
Dedicated acceleration for maximum throughput and low latency
- Massive parallel processing
- Ultra-low latency
- Energy efficient
Binary SDK
Embed high-performance compression directly in your applications
- C/Fortran & Python APIs
- Cross-platform support
- Minimal dependencies
Sealed Gateway Appliance
Rack-ready hardware for fast, secure on-prem deployments
- Turnkey deployment
- Enterprise-grade security
- Scalable throughput
Licensing & Terms
Evaluation (PoC): NDA with time-limited licenses and benchmark tooling. Production: licensed by throughput (GB/s) or per gateway/appliance, with offline activation available for on-prem deployments.
Operating modes: offline / air-gapped operation is supported for the on-prem SPIRAL-BOX, SDK, Sealed Gateway Appliance, and FPGA Offload. The cloud-managed service requires internet connectivity. Minimum system: 1 CPU. Scaling: near-linear with additional cores or cards, subject to I/O and bandwidth.
Our Team
Eduin Serna
Founder · Physicist & Computer Scientist
Andres H Serna
Founder · Electronic Engineer & Remote Sensing Scientist
Our team brings over 15 years of expertise spanning quantum physics, cryptography, computer science, electrical engineering, and remote sensing. We have worked with NASA, Fermilab, CERN, and top universities, advancing quantum key distribution, high-performance computing, and pioneering patented lossless data compression. Beyond this, our experience with LiDAR and satellite data analysis has fueled innovation across diverse sectors. With more than 27 publications in leading journals—including Nature, Science, Physical Review Letters, and PNAS—and over 3,800 citations, we also serve as Artificial Intelligence advisors to the State of Maryland. Guided by this foundation, we continue to shape the future of communication, data storage, and digital innovation.