INVESTOR & PRODUCT PRESENTATION · 22 OCTOBER 2026
GLYX Archive.
The product and its evidence.
Benefits, recorded results and the proposed licensing route. Each comparison keeps its workload and measurement boundary.
UK patent application GB2624952.4 · Pending
Slide 1 of 12
Use the arrow buttons or focus the presentation and press the left or right arrow key. The PDF includes all 12 slides. Read the full text below.
ACCESSIBLE TEXT VERSION
Read the presentation.
Recorded studies support scoped evaluations. Production delivery and commercial terms require agreement.
01. GLYX Archive
UK patent application GB2624952.4 (pending). Recorded results and a proposed commercial route.
02. The value of a growing shared archive
Related files and versions often repeat useful structure. Retaining it together creates an opportunity to reduce data overhead.
People and teams
Keep compatible libraries compact and restore selected files.
Business services
Evaluate storage, transfer and access costs on the same data.
AI applications
Retain exact tool data, structured histories and verified artifacts.
Devices and applications
Explore compact persistence within an agreed integration.
Prospective applications. Customer demand and deployment economics require evaluation.
03. The archive workflow
A simple archive experience, with exact restoration as the acceptance condition.
Choose content
Add supported files or folders to a shared archive. Keep originals while checking the result.
Review the saving
Compare source size with the complete archive and restoration record.
Restore selected files
Choose content and a destination. Other archived content can remain compact.
Verify the result
Check exact restored bytes and retain a recovery copy under your own policy.
04. 742 files restored exactly
A private candidate compared the same 1.651 GB of source content with a same-content ZIP.
46.8%
smaller than the same-content ZIP
1,490,471,935 bytes versus 793,564,456 bytes.
All 742 files restored exactly.
| Same-content ZIP | 1490.471935 |
|---|---|
| GLYX + restoration record | 793.564456 |
Runtime and installation are additional. Counting both conservatively gives about 845.4 MB, or 43.3% below ZIP.
05. Reuse can improve as the library grows
More compatible content can reduce the marginal storage needed for later additions.
One retained library
Related content can share applicable structure within the archive.
Exact changes remain
Unique information and reconstruction dependencies still count.
Selected access
Restore the requested content without materialising the whole library.
Measure each workload
File size alone does not guarantee higher compression or faster restoration.
Encrypted, already-compressed and unique data can behave differently.
06. Faster complete archive operations
Six authored workloads. Median complete-operation time, six repetitions per arm.
| Workload | Existing | Candidate | Time change |
|---|---|---|---|
| Mixed content | 351.729 ms | 299.781 ms | 14.77% less |
| Repeated content | 1,155.853 ms | 894.071 ms | 22.65% less |
| Unique content | 338.421 ms | 288.503 ms | 14.75% less |
| Structured records | 577.064 ms | 582.620 ms | 0.96% more |
| Multiple sections | 1,070.307 ms | 1,034.495 ms | 3.35% less |
| Append operation | 435.288 ms | 357.410 ms | 17.89% less |
4/6 workloads met the gate. All 1,656 exact-output checks passed; archive sizes unchanged. Peak child memory was about 129 MB versus about 34 MB on several baseline workloads.
07. Storage and time in one component study
One synthetic versioned-data workload. 96 formation versions and 1,536 fresh versions per arm.
| Implementation | Pipeline | Logical data | Allocated space |
|---|---|---|---|
| Archive/state component | 135.98 s | 32.51 MB | 246.02 MB |
| Trained Zstandard | 82.81 s | 76.09 MB | 289.41 MB |
| Borg + compaction | 918.48 s | 52.31 MB | 445.78 MB |
57.3% fewer logical bytes than trained Zstandard
37.9% fewer than compacted Borg. All 9,216 fresh full and selected-range reconstruction checks passed.
Separate component study. Zstandard was faster. Logical data, allocated space and installation have different boundaries.
08. The product preserves exactness
The useful promise is an exact retained record with a clear storage and recovery boundary.
Growing shared archives
Reuse compatible structure across files and versions.
Large-file streaming
Process files in sections with a bounded working set.
Exact selected restoration
Reconstruct and verify the content requested.
Visible savings accounting
Show archive and restoration costs, with backups and application storage separately.
Capacity depends on available storage and the qualified platform.
09. Licensing follows the deployment
Non-exclusive usage and distribution rights, with BBF retaining technology ownership.
People and small teams
Proposed device or team subscription. Set pricing for the qualified release.
Business and AI services
Proposed 50% share of verified net eligible savings, or annual capacity bands.
Application and device makers
Agree deployment rights, capacity bands and OEM terms.
Coordinated BBF stack
Use one cost ledger so the same saving enters the commercial calculation once.
Commercial proposals. No live checkout, automatic billing or accepted customer terms are implied.
10. A focused route to customer release
Private engine and desktop-candidate evidence supports the next delivery steps.
Current evidence
Exact archive restoration and defined storage comparisons.
Platform acceptance
Windows archive/CLI: 92 scoped checks. Full workflow qualification remains in progress.
Distribution
Qualify installation, signing, updates and recovery on the target host.
Customer evaluation
Measure the actual data, access pattern and complete operating economics.
Hardware, operating systems and applications receive their own acceptance checks.
11. An application within a product family
UK patent application GB2624952.4 is pending. The company licenses defined product capabilities.
GLYX Archive
Compact exact persistence and selected restoration.
Seed State OS
Qualified reconstruction of required working state.
TokenHoloFold
Supported local execution inside wider AI workflows.
Combined delivery
Qualify the full integration and count shared financial benefits once.
Applications are pending. Combined performance and customer economics require their own evaluation.
12. The next commercial milestones
Turn recorded capability
into qualified deployment.
Priorities for a product programme:
01
Complete target-host delivery and recovery acceptance.
02
Run a scoped customer evaluation on real retained data.
03
Agree licensing, support and repeatable distribution.
Discuss your dataset: www.beyondbruteforce.com | hello@beyondbruteforce.com
UK patent application GB2624952.4 is pending. Read the evidence and comparison boundaries →