ITAR / EAR · DETERMINISTIC CLASSIFICATION ENGINE
Regulus replaces hours of manual ITAR/EAR research with a deterministic logic engine. The same technical parameters always produce the same defensible result — every time.
Two senior GTC specialists analyzing the same hardware under the same USML category regularly produce different rationales. Neither is wrong — but only one of them will survive a DDTC audit five years later.
PRIMARY DRIVER OF REGULATORY EXPOSUREDDTC and BIS retrospective audits can cover thousands of parts across years. Without a traceable, reproducible record of how every determination was made, companies cannot defend themselves — and legal fees alone run into millions.
$2–8M TYPICAL LOOKBACK COSTInconsistently phrased DECCS and SNAP-R submissions trigger formal Requests for Information — forcing costly technical deep-dives that could have been prevented by capturing the right data at intake.
AVG 6–18 WEEK RFI DELAY"I built the first version of Regulus internally to solve these exact problems. The results were immediate: a 90% reduction in classification time and a drastic decrease in disagreement between specialists."
— FOUNDER, FORMER GTC TECHNICAL SPECIALIST, L3HARRIS / AEROJET ROCKETDYNE TEAM| Manual / Spreadsheet | ONESOURCE / KYG AI | REGULUS | |
|---|---|---|---|
| Classification method | Expert judgment per specialist | Probabilistic text-matching (confidence scores) | Deterministic Boolean gates — same input, same output, always |
| Rationale consistency | ✗ Varies by specialist | ✗ Varies by model version | ✓ 100% standardized across all users |
| Audit trail | ✗ Email chains, spreadsheets | ✗ Log files, not legally defensible | ✓ Immutable ledger with SHA-256 gate-level hashes |
| §120.41 SD analysis | Manual lookup, prone to error | Not explicitly modeled | ✓ Full CATCH / RELEASE / Same-As logic built in |
| RFI pre-emption | ✗ Post-hoc scramble | ✗ Not addressed | ✓ Mandatory DDTC/BIS RFI fields captured at intake |
| Human accountability | Implicit — hard to trace | None — AI made the call | ✓ Every gate timestamped and specialist-approved |
| Time to rationale | 2–6 hours per part | Minutes, but still needs specialist validation | ✓ <5 min specialist review after automated research |
| DECCS / SNAP-R ready | ✗ Reformat manually | Partial | ✓ Structured export mapped to agency portal fields |
An internal prototype at a major aerospace Prime reduced the research-to-draft cycle from several hours to under 10 minutes per classification — without compromising legal defensibility.
The deterministic gate model eliminated inter-specialist variance on identical hardware. The same part number classified by two different specialists produces the same rationale, every time.
Coverage across USML IV (Propulsion), VII (Ground Vehicles), VIII (Aircraft/UAV), XI/XII (Electronics/Sensors), XV (Space), and the full EAR CCL 600-series and dual-use stack.
Stop spending six hours researching a single classification. Regulus automates the research phase and presents every gate for your review. You approve the logic — in under five minutes.
See every session across your team in real time. Review the assembled rationale before signing off. Flag classifications that need further work — with a permanent audit record of why.
When DDTC or BIS comes asking, you hand them a timestamped, hash-verified record of every gate decision, every specialist approval, and every technical parameter evaluated — down to the second.
Shipping a custom bracket for a military UAV program? Regulus generates the same quality rationale a large Prime would produce — without requiring deep regulatory expertise.
Regulus is in early access for select defense prime contractors and Tier 2 suppliers. Request access and we'll be in touch within 48 hours.
No spam. We contact you directly to schedule a 30-minute demo.