Navy SBIRPhase I
TRL3-4Multi-domain AI/ML R&D from Inception to Commercialization.
Over a decade of deep-tech R&D for the Department of Energy and the Department of Defense.
Click here to see all 26 awards
CyberNeuro-RT
Real-time network defense that learns your network instead of memorizing someone else's.
Read more →
Discovium
Physics-first discovery of materials that stay strong at extreme temperature.
Read more →
E-RECOV
Recovers gold, copper and other metals of value from electronic waste electrochemically, at room temperature.
Read more →- Inception
- Research
- Development
- Integration
- Commercialization
Products
AI-native network defense, physics-guided materials discovery, and green-tech metal recovery from electronic waste.
01
Agentic Network Defense Ecosystem
Cybersecurity - Penetration Testing - Source Code Vulnerability Detection
Flagship · Network defense
CyberNeuro-RT (CNRT)
Real-Time Intelligence > Anywhere Deployment > Stronger Defense.
- Speed: Detect threats while intervention still matters.
- Scalability: Protect distributed environments through a flexible architecture.
- Integration: Strengthen existing cybersecurity investments with an additional layer of AI-driven intelligence.
Detect Threats in Real TimeCNRT identifies emerging threats as network activity unfolds, enabling faster, AI-driven detection and response before attacks escalate.
Enterprise-Scale Security, Edge-to-CloudDeploy intelligent threat detection across edge, on-premises, hybrid, and cloud environments through one scalable security architecture. CNRT provides flexible, distributed protection without compromising operational visibility.
Throughput1.41Gb/s
Anomaly0.02Baseline
Sensor draw4.1Watts
Packet analysis rate
141800
Event stream
ReconOutbound volume within learned envelopeCLEARED
ModelModel checkpoint rolled forwardUPDATED
EdgeEdge sensor reported inSYNCED
Monitoring
Mass file-encryption pattern
Contained
Segment B · host 10.4.22.8
Detected13:42:00.774
Files touched1,284
Encrypted0
ActionIsolated
Why it was flagged
A single host began reading and rewriting files across three shares at a rate no user on this network has ever produced. No signature was involved. The host was isolated before any file was written back in an encrypted state.
On mass-encryption patternIsolate host, lock shares
On credential reuseAlert analyst, no automatic action
On unknown destinationRate-limit, then alert
Analyst overrideAlways available
Strengthen Your Defenses Without Replacing ThemEnhance your existing UTM (Unified Threat Management) and NGFW (Next-Generation Firewall) security investments, not replace them. CNRT adds an AI-driven detection and intelligence layer that complements firewalls, endpoint protection, and traditional security infrastructure, helping identify sophisticated threats and anomalous activity that conventional defenses may overlook.
Autonomous penetration testing
CNRT Vanguard
An autonomous pentester. It plans and runs the full kill chain against systems you have authorized, from reconnaissance to domain admin, and shows you every step.
A kill chain, not a bag of scripts66 attack techniques, each mapped to MITRE ATT&CK, chained by playbooks from first packet to domain admin.
Reason, act, observe, and you watchThe agent runs real tools, reads what comes back and decides the next move, with every step streamed to a live console.
Fail-closed by defaultA run reaches nothing until you authorize its scope, and that boundary is enforced outside the agent.
Target3 hostsIn signed scope
Paths found0Not yet run
EgressClosedOpens to scope only
Kill chain
Awaiting run
ReconIdle
Initial accessIdle
ExecutionIdle
Priv-escIdle
LateralIdle
Domain adminIdle
Event stream
ScopeEngagement 214 countersignedVERIFIED
EgressClosed until scope is attestedFAIL-CLOSED
Authorized, idle
1
vpn-gw.customer.example
REACHABLE
Credential reuse from a service account gave a foothold on the gateway. Reproduction steps and evidence captured.
2
app-01.customer.example
PARTIAL
Escalation attempted from the foothold. Host policy held, so the path stops here.
Authorized byHead of Information Security
Window14 days from countersignature
ReferenceENG-214-A
Systems named in the authorization
app-01.customer.example
IN SCOPE
vpn-gw.customer.example
IN SCOPE
everything else
OUT OF SCOPE
DevSecOps · AI vulnerability scanning
CNRT CodeLens
An AI vulnerability scanner for source code. It finds the vulnerabilities an attacker could actually reach and puts that short list in front of the people who fix them.
Ranked by what is reachableTwo exploitable findings beat two hundred theoretical ones nobody will work through.
Reads a repository like a reviewerFollows how the code is actually called rather than pattern-matching on syntax.
Fits the pipeline you already runNo new process to adopt and no separate system to keep in step.
auth/ session.c token.cnet/ parse.c frame.c
auth/session.c
208
size_t need = hdr->len + sizeof(tag_t);209
if (need > cap) {210
cap = need;211
}212
213
memcpy(buf, hdr->body, hdr->len);214
buf[hdr->len] = 0;215
216
return finalize(buf, need);217
}auth/token.c
41
int token_verify(const tok_t *t, time_t now) {42
if (!t || t->len == 0)43
return TOK_EMPTY;44
if (t->exp < now)45
return TOK_EXPIRED;46
47
uint8_t mac[32];48
hmac_sha256(key, t->body, t->len, mac);49
return ct_equal(mac, t->mac, 32) ? TOK_OK : TOK_BAD;50
}net/parse.c
88
size_t parse_header(const uint8_t *p, size_t n, hdr_t *h) {89
if (n < HDR_MIN)90
return 0;91
h->type = p[0];92
h->len = read_u16(p + 1);93
h->body = p + HDR_MIN;94
95
if (h->len > n - HDR_MIN)96
h->len = n - HDR_MIN;97
return HDR_MIN + h->len;net/frame.c
130
int frame_next(stream_t *s, frame_t *f) {131
hdr_t h;132
size_t used = parse_header(s->cur, s->left, &h);133
if (used == 0)134
return FRAME_SHORT;135
136
f->kind = h.type;137
f->data = h.body;138
s->cur += used; s->left -= used;139
return FRAME_OK;
Reachable
Unchecked bounds on copy path
Length comes from the caller and is used to write before it is compared against the destination size.
Nine stages, resumable
1Index repositoryqueued
2Resolve dependenciesqueued
3Build call graphqueued
4Multi-lens detectionqueued
5Cluster candidatesqueued
6Verify findingsqueued
7Rank by reachqueued
8Draft remediationqueued
9Export reportqueued
auth/session.c:213
Proposed change, for your review
212
213
- memcpy(buf, hdr->body, hdr->len);213
+ if (hdr->len >= cap) return E_RANGE;214
+ memcpy(buf, hdr->body, hdr->len);ApplySkip
Export as MarkdownExport as PDF
02
Materials Discovery
Transforming trial-and-error experimentation into predictive, physics-guided optimization
Discovium
Engineering the Next Generation of Extreme-Performance Materials.
An autonomous, physics-driven AI platform that predicts, optimizes, and validates advanced materials based on three critical factors: strength, durability, and predictive confidence, accelerating breakthrough discoveries for the world's most demanding environments.
1M+first-principles calculations completed
11,000+High-Entropy Alloy (HEA) compositions fully evaluated
3dimensions scored for every candidate
24/7autonomous discovery loop
03
Metal Recovery
Green-Tech recycling of electronic waste.
E-RECOV
AI-Driven, Characterization-Controlled, Continuous, Multi-Metal Recovery from E-Scrap and Permanent Magnets
Developed at the Critical Materials Institute and incubated at Idaho National Laboratory, E-RECOV's preprocessing, separation, and characterization pipeline performs tests to determine e-scrap particle sizing to maximize material extraction, calibrates the flow physics of electrochemical & hydrometallurgical processes of recycling e-scrap in real time, and uses 75% less chemical reagent, 30% less energy, and 25% less water intake and wastewater generation to do all of this at room temperature.
Gold96%
Recovered from feedCopper91%
Recovered from feedReagent used25%
Against conventional baseline1 Leach2 Separate
3 Elute4 Recover
Domains
26 Federal contracts awarded since 2019
Concurrent NASA, Air Force, and Navy projects, from research through commercialization.
Recent Projects
NAVY
Imaging
Hyperspectral detection of uncrewed aircraft
Detects and classifies small uncrewed aircraft from hyperspectral imagery, using spectral signature rather than shape, so targets that defeat conventional optical tracking are still resolved.
Learn More →
NAVY
Synthetic aperture radar
Real-time synthetic aperture radar simulation
Generates radar imagery as fast as it would arrive in the field, so detection and classification models can be trained against scenarios that are impractical or far too expensive to actually fly.
Learn More →
DOE
Materials
Electrochemical recovery of metals from electronic waste
Recovers gold, copper and other constituents of value from electronic waste electrochemically and at room temperature, with no smelter and none of the emissions handling one needs. Commercialized as E-RECOV.
Learn More →Contract showcase
Recent SBIR / STTR Awards
18 Awards6 AgenciesTRL 2–6 Readiness13 Partners
| Sponsor | Research topic | TRL | Specialization area | Collaborating partners |
|---|---|---|---|---|
| Navy SBIRPhase I | Hyperspectral image detection of UAVs | 3-4 | AI/ML & Hyperspectral imaging | Lockheed Martin & Bodkin Design & Engineering |
| MDA STTRPhase I | Ground-based hypersonic threat detection using neuromorphic processing | 3 | AI/ML, Hypersonics & Neuromorphic computing | Lockheed Martin & University of Florida |
| DHS SBIRPhase I | Contraband detection using hyperspectral imaging and neuromorphic computing | 3 | AI/ML, Hyperspectral imaging, & Neuromorphic computing | Lockheed Martin & Bodkin Design & Engineering |
| Army SBIRPhase I | Multi-sensor object detection and tracking | 3 | AI/ML & Computer Vision | Dr. Raktim Bhattacharya |
| Air Force SBIRPhase II | Imaging spectropyrometer for hypersonic thermal protection system characterization | 4 | AI/ML & Hyperspectral imaging | Bodkin Design and Engineering (prime) |
| Navy SBIRPhase I | Multi-sensor UAV for corrosion evaluation and characterization | 3 | AI/ML & Hyperspectral imaging | Bodkin Design and Engineering (prime) |
| Navy SBIRPhase I | Source code vulnerability detection | 4 | AI/ML & Cybersecurity | Lockheed Martin & University of Wisconsin |
| DOE SBIRPhase II | HPC-scale cyber-threat detection using neuromorphic computing | 4 | AI/ML, Cybersecurity & Neuromorphic computing | Pennsylvania State University |
| MDA SBIRPhase I | Cyber-threat detection | 4 | AI/ML & Cybersecurity | Lockheed Martin |
| Army STTRPhase I | Automatic threat detection using metamaterial optics | 3 | ATR, metamaterials, AI/ML | California Institute of Technology |
| Navy SBIRPhase I | Smart image recognition sensor using metamaterial optics | 3 | ATR, metamaterials, AI/ML | California Institute of Technology |
| Navy SBIRPhase I | Metamaterial enhanced micromirror surfaces for enhanced IR beam control | 3 | Metamaterials, Infrared imaging | Duke University |
| DOE STTRPhase IIA | Electrochemical Recycling Electronic Constituents of Value (E-RECOV) | 6 | Precious metals, Base metals, and REE extraction from E-Waste | Idaho National Lab & Colt Refining and Recycling |
| Navy SBIRPhase I | ML database to guide development of low flammability polymer matrix composites | 3 | Materials sciences, AI/ML | Dr. Richard E. Lyon |
| Navy SBIRPhase II | Real-time SAR simulator | 4 | SAR simulation & Hypersonics | Lockheed Martin & Dr. Andrew Willis |
| Air Force SBIRPhase I | Context-Aware RF EM Surveying for Exploiting Signals of Opportunity | 3-4 | RF surveying (detection and localization) | Lockheed Martin & Dr. Andrew Willis |
| Navy SBIRPhase I | Next generation tool sets for weapons separation evaluations | 3 | Modeling and simulation (aerodynamics) | Dr. Raktim Bhattacharya |
| DOE SBIRPhase I | AI Based Forecasting Models for Amine and Degradation Product Emissions | 2-3 | Computational Chemistry, AI/ML | Electric Power Research Institute |
MDA STTRPhase I
TRL3Ground-based hypersonic threat detection using neuromorphic processing
DHS SBIRPhase I
TRL3Contraband detection using hyperspectral imaging and neuromorphic computing
Army SBIRPhase I
TRL3Multi-sensor object detection and tracking
Air Force SBIRPhase II
TRL4Imaging spectropyrometer for hypersonic thermal protection system characterization
Navy SBIRPhase I
TRL3Multi-sensor UAV for corrosion evaluation and characterization
Navy SBIRPhase I
TRL4Source code vulnerability detection
DOE SBIRPhase II
TRL4HPC-scale cyber-threat detection using neuromorphic computing
MDA SBIRPhase I
TRL4Cyber-threat detection
Army STTRPhase I
TRL3Automatic threat detection using metamaterial optics
Navy SBIRPhase I
TRL3Smart image recognition sensor using metamaterial optics
Navy SBIRPhase I
TRL3Metamaterial enhanced micromirror surfaces for enhanced IR beam control
DOE STTRPhase IIA
TRL6Electrochemical Recycling Electronic Constituents of Value (E-RECOV)
Navy SBIRPhase I
TRL3ML database to guide development of low flammability polymer matrix composites
Navy SBIRPhase II
TRL4Real-time SAR simulator
Air Force SBIRPhase I
TRL3-4Context-Aware RF EM Surveying for Exploiting Signals of Opportunity
Navy SBIRPhase I
TRL3Next generation tool sets for weapons separation evaluations
DOE SBIRPhase I
TRL2-3AI Based Forecasting Models for Amine and Degradation Product Emissions
No awards match that search.
Our Joint Venture
Vulcan Ventura
Three programs in RF sensing, navigation and radar autonomy.
Air Force
RF Emitter Location
A compact onboard sensing package that locates and identifies radio-frequency emitters from unmanned aircraft, holding its accuracy in cluttered signal environments.
Navy
RF-Referenced Navigation
A compact phased-array system that uses existing radio transmissions as navigation references, giving fast-moving vehicles an independent source of position data.
NASA
Compact Radar for Autonomy
A low-cost radar sensing approach that gives autonomous vehicles perception and navigation where cameras and lidar fall short, in darkness, dust and poor light.
Partnerships, Certifications & Consortiums
Global presence
A multifaceted operation. In government work you cannot be a single-focused company, and you cannot be in one place either.
Silicon ValleyCarson CityVancouverDubaiTokyo
Silicon Valley, California
Quantum Ventura Inc.
Carson City, Nevada
Chemical engineering
Vancouver, British Columbia
Meta-optics and metamaterials
Dubai, U.A.E.
Quantum Guru Division
Tokyo, Japan
Distribution and support
Tell us what you're working on.
Our team replies within two working days, wherever you are.