"We Shrink the Search Space"

Seed Round — Confidential
150M+
OPS/SEC
72 GPU
Cluster
1078→103
Reduction
$50M
Valuation
Scroll
01 — Mission

Make the impossible
searchable.

Gondor builds quantum-inspired search-space reduction engines that collapse computationally infeasible problem domains into tractable solutions — transforming months of brute-force computation into seconds of precision.

Medical Discovery

Protein folding, drug interaction mapping, genomic search

Defense & Security

Signal intelligence, threat detection, cryptographic analysis

Cryptography

Key-space reduction, entropy analysis, cipher validation

Financial Security

Fraud detection, risk modeling, anomaly identification

Complex Optimization

Supply chain, logistics, combinatorial scheduling

Enterprise AI

Model pruning, inference acceleration, data deduplication

The Opportunity

What would you do?

Bitcoin Core wallet showing 12,600 BTC locked

You’re locked out of this wallet.
What do you do?

3–4M
Lost BTC
$300B+
Locked Value
20%
Of All BTC Supply
02 — Problem

Brute force
is dead.

The world's most critical computational problems exist in search spaces so vast that no amount of hardware can solve them by exhaustive enumeration.

1078

Impossible Scale

Typical cryptographic search spaces exceed the number of atoms in the observable universe

99.97%

Wasted Compute

Nearly all GPU cycles in brute-force approaches evaluate candidates that could have been mathematically eliminated

Months→∞

Time to Solution

Even state-of-the-art GPU clusters cannot meaningfully dent these search spaces within operational timeframes

$0 ROI

Negative Returns

Organizations spend millions on compute infrastructure with no guarantee of finding solutions

03 — Solution

Quantum collapse of
search space.

1075×
Search Space Reduction
PHASE 1
Ballistics
Discover password length L∈[8,20] via CHEAP32 oracle on 72 GPUs
PHASE 2
Spike Maps
Refine character positions via CV sealed entropy oracle + SPRT pruning
PHASE 3
Domain Pruning
Narrow character domains per position via evidence-weighted alpha
PHASE 4
Full Verify
AES-256-CBC verification on GB200 cluster — password found

Before Gondor

Search space: 1078 candidates
Time: Billions of years
Cost: Effectively infinite
Success: Not possible

After Gondor

Search space: 103 candidates
Time: Hours to weeks
Cost: $250K–$500K compute
Success: Probabilistic w/ SPRT bounds
04 — Technology

The Spider
Engine.

CV Qumode Signal Models

Continuous-variable quantum-inspired signal decomposition for structural pattern extraction

QUMODE CV-QNN

Entropy Pruning

Sequential probability ratio testing for provable elimination of infeasible search regions

SPRT/Wald Shannon

Probabilistic Scoring

Bayesian candidate ranking with adaptive confidence thresholds and real-time recalibration

Bayesian MCMC

DAG Feedback Loops

Directed acyclic graph orchestration enabling iterative refinement with convergence guarantees

DAG Feedback

Custom CUDA Kernels

Hand-optimized sm_90 GPU kernels for GB200 NVL72, maximizing throughput on NVIDIA's latest architecture

GB200 sm_90 NVL72
150M OPS/SEC
on NVIDIA NVL72 Cluster
05 — Competitive Advantage

Why no one else
can do this.

Four-Phase Pruning Circuit

Ballistics → Spike Maps → Domain Pruning → Full Verify. Each phase eliminates orders of magnitude from the search space before the next begins.

Quantum Entropy Oracle

Sealed CV qumode oracle returns only entropy_bits — no raw candidates leak. Gaussian/Fock backends guide pruning with quantum-inspired signal decomposition.

SPRT Statistical Engine

Wald Sequential Probability Ratio Test drives every tile decision: Accept, Reject, or Continue. FDR ledger provides SHA256-chained cryptographic audit trail.

Capability Gondor Brute Force Heuristic
Quantum-guided pruning
1075× search reduction
Password length detection
SPRT accept/reject/continue
72-GPU custom CUDA kernels
Cryptographic audit trail
06 — Market

$340B+ addressable
market.

$340B
TAM
$40B
SAM
$1.2B
SOM
$300B+
Locked Crypto
3-4M lost BTC, 11-20% of total supply inaccessible
$25B
Government & Defense
328K seized BTC, SIGINT, cryptanalysis tools
$13B
Digital Forensics
Password recovery, device exploitation, eDiscovery
$6B
Security Compute
GPU-as-a-Service for cryptographic workloads
07 — Business Model

We recover
locked value.

$250K retainer to engage. 50/50 split on recovered assets. Retainer reimbursed on successful recovery.

50 / 50

Crypto Asset Recovery

$250K retainer to begin. Half of all recovered value. Retainer reimbursed upon successful wallet opening. One whale wallet = one year of operations.

$500K–$10M+

Government & Law Enforcement

IDIQ contracts, OTA agreements, and SBIR programs for seized crypto access and national security cryptanalysis.

Usage

Compute API

Pay-per-reduction cloud API metered by compute-seconds. 905M effective ops/sec on GB200 cluster.

The Team

Built by
operators.

Jay Westerdal

CEO

15+ years cybersecurity & engineering. Founder of DomainTools (acquired). Serial entrepreneur in security infrastructure.

Sky Trotter

CTO

Post-quantum cryptography, blockchain security, and GPU kernel engineering. Architect of the Spider Engine and CUDA pipeline.

Advisors
Joshua Goldbard

Founder of MobileCoin — the privacy-first cryptocurrency integrated into Signal Messenger.

Matt Cantieri

Founder of Hyperdrive (top DeFi on Hyperliquid). Managing Partner, Three Sigma Capital. Ex-Microsoft Director.

08 — Traction

Built and
validated.

93+
White Papers Published
6000+
Technical Audits
80+
Engine Components
30
PhD-Level Chapters
COMPLETED
Core Engine Built
Spider Engine v1 with 80+ components, CUDA kernels, and full pipeline validated on GB200
Q3 2026
Seed Close + First Contracts
Close $5M seed round, sign first government and enterprise pilot agreements
Q4 2026
Cloud API Launch
Compute-as-a-Service platform live with metered pricing and self-serve onboarding
2027
Scale + Series A
Expand to 10+ enterprise customers, demonstrate $4M ARR, raise Series A
09 — Financials

Path to
$175M+ ARR.

$2.8M
2026
$17.5M
2027
$57M
2028
$112M
2029
$175M
2030
75%
Gross Margin
135%
Net Revenue Retention
<6mo
Payback Period
10 — The Ask

Raising $5 Million
Seed.

$5M
Seed Round — Priced Equity
Equity 10%
Pre-Money Valuation $50M
Post-Money Valuation $55M
Instrument Priced Equity
Target Close Q3 2026

Use of Funds

40%
25%
20%
15%
Engineering $2M
GPU Infrastructure $1.25M
Go-to-Market $1M
Operations $0.75M

Key Milestones

  • First 3 enterprise contracts signed
  • Cloud API public launch
  • 72-GPU cluster fully operational
  • $4M ARR run-rate by Q4 2027
  • Series A readiness

"We Shrink the Search Space"

The world's hardest problems aren't waiting for faster GPUs.
They're waiting for smarter search.

$5,000,000
for 10% equity