AutoeCTD Inc. · Federally incorporated Vol. 1 · Issue 01 · 2026
It's time to reinvent the wheel.

Regulatory Submissions
built by Professionals.
Accelerated by AI.

AutoeCTD is the regulatory submission platform built by RA professionals, for RA professionals — with AI doing the repetitive work and your reviewers making every decision that matters. Authoring, labeling, eCTD, and deficiency responses, in one suite.

90%+
Document classification accuracy goal with the full three-layer engine.
6×
Products in the suite. Authoring, labeling, XML, eCTD, deficiency responses, and quality lifecycle.
3
Major pharmaceutical markets — built for jurisdictions like the FDA, EMA, and Health Canada, and any agency that accepts the eCTD format.
0
Legacy compromises. AI-native architecture, not bolted on.
app.autoectd.com / submissions

Active submissions

+ New
SAMPLE-DRUG-A
NDA-000001 · Module 1–5 · 247 documents
100% Ready
SAMPLE-DRUG-B
IND-000002 · Module 2–5 · 156 documents
78% Reviewing
SAMPLE-DRUG-C
NDA-000003 · Module 4 · 89 documents
42% Classifying

ILLUSTRATIVE INTERFACE · DRUG NAMES AND APPLICATION NUMBERS ARE FICTITIOUS

§ 01 — The thesis

Every regulatory submission tool on the market today was designed before modern AI existed. The result is software that asks Regulatory Affairs professionals to do the AI's job — manually classifying, naming, and wiring up thousands of documents, one drag at a time.

We built AutoeCTD on a simple premise: the work of compiling an eCTD submission for jurisdictions like the FDA, Health Canada, or the EMA — or any agency that accepts the format — shouldn't take weeks of a senior reviewer's time. The intelligence to do most of it well already exists — it just hasn't been put inside the tools your team uses.

We don't replace Regulatory Affairs judgment. We give it eight hours of its workday back.

Our customers are mid-size pharma teams with multiple marketed products, high submission volume, and a sharp answer to the question "why are we still naming PDFs by hand?" If that's you, keep reading.

§ 02 — The AutoeRegulatory Suite

One platform, six workflows. From document authoring to post-approval lifecycle.

AutoeCTD

eCTD submission packaging for jurisdictions like the FDA, EMA, and Health Canada — and any agency that accepts the format.

Drop a folder.
Get a valid sequence.

Upload your raw documents. Our classification engine assigns them to the correct ICH M4 Module 1–5 sections, auto-fills the regional cover form, computes MD5 checksums, and writes the eCTD XML backbone. You review. We package — for whichever agency you're filing with.

  • AI document classification into ICH M4 Module 1–5 sections
  • Regional cover form auto-population (FDA Form 356h, Health Canada equivalents, more)
  • eCTD v3.2.2 XML backbone with regional adapters
  • MD5 integrity checksums for every leaf
  • SharePoint package delivery
In pilot Multi-jurisdiction · ICH M4
AutoeXML

Regulatory XML automation — FDA SPL and Health Canada XML PM.

From Word draft
to gateway-ready XML.

Drop in a labeling document. AutoeXML extracts the sections, classifies them against jurisdiction-specific code libraries (FDA SPL / Health Canada SPM 2020), validates against official vocabularies, and emits schema-valid XML — ready for ESG transmission.

  • FDA SPL v3 + Health Canada SPM 2020 adapters
  • Three-layer classification with LOINC and HPFB code libraries
  • Live validation against official FDA and HC vocabularies
  • OTC, REMS, establishment registration, drug listing
  • e-signature workflow with append-only audit trail
In pilot FDA SPL · HC SPM 2020
AutoeDEF

Deficiency response intelligence.

Never write the same response twice.

AutoeDEF reads incoming agency correspondence, extracts each deficiency point, and drafts a response grounded in your own historical submissions. Track what was accepted, what was rejected, and benchmark your performance by category and agency.

  • Automated deficiency extraction from agency correspondence
  • Vector search across your historical response library
  • Grounded draft generation with source citation
  • Outcome tracking — Accepted / Rejected / Partial
  • Response effectiveness benchmarking
In pilot Multi-agency
AutoeDOC

Regulatory document authoring — Module 2 summaries, prescribing information, and submission-ready drafts.

Draft from source.
Not from a blank page.

AutoeDOC drafts the documents that go into a regulatory submission — Quality Overall Summaries, Clinical and Nonclinical Overviews, prescribing information, risk management plans — using your source data and prior-approved precedent. Your senior writers review, refine, and approve. The blank-page problem disappears.

  • Module 2 summary drafting from CMC, nonclinical, and clinical data
  • Product Monograph and prescribing information generation
  • Legacy PM conversion to the 2020 Master Template
  • Source-grounded drafting with citation back to underlying study reports
  • Append-only audit trail of every AI suggestion and human edit
In pilot FDA · Health Canada · ICH M4
AutoeLABEL

Branded and generic drug labeling — creation, updates, and reference-product alignment.

One label engine.
Brand and generic.

AutoeLABEL creates new drug labels and keeps generic labels aligned with the Canadian Reference Product. When the brand updates a safety section, the generic label updates with a drafted change for your team to review and submit. No more manual diffing across hundreds of generic SKUs.

  • Branded label authoring from prescribing information and CMC data
  • Generic label maintenance with automatic CRP alignment
  • Brand safety update detection across reference product sources
  • Side-by-side diff for review and submission of label changes
  • Bilingual handling for Health Canada English/French requirements
In active development FDA · Health Canada
AutoeQA

Quality lifecycle management — change controls, deviations, and CAPA workflows.

Quality events.
Connected to submissions.

AutoeQA handles change controls, deviation investigations, and CAPA workflows — connected directly to the regulatory submission pipeline. When a manufacturing change is approved, the downstream label and submission updates are drafted automatically. Quality and Regulatory, finally talking to each other.

  • Change control workflow with regulatory impact assessment
  • Deviation and investigation case management
  • CAPA tracking with effectiveness review
  • Auto-trigger of downstream label and submission updates
  • 21 CFR Part 11 audit trail across the quality lifecycle
On the roadmap 2026–2027
§ 03 — The classification engine

Three layers of intelligence. One auditable decision per document.

How it works

Cheap signals first. Expensive AI last.

Most AI classification tools throw every document at a large language model and call it a day. That's slow, expensive, and impossible to audit.

AutoeCTD layers four cheap deterministic signals first — filename patterns, version inference, prior-sequence cross-reference, file-size outliers — then escalates only the residual through a manufacturer-aware prompt library, and finally a generic AI fallback. Every decision is logged with the layer it came from and a confidence score.

90%+
Combined accuracy target with all three layers and a mature Scan & Learn library.
Layer 01
Deterministic No AI

Filename pattern match · version inference · prior-sequence cross-reference · file-size outlier check. Handles roughly half of every submission at zero token cost.

Layer 02
Prompt library Manufacturer-aware

Manufacturer-specific prompts plus first-page text. Confidence stacks across DS Manufacturer, DP Manufacturer, Sponsor, and CRO. Scan & Learn auto-populates the library from each new submission record.

Layer 03
Generic AI fallback GPT-4o · Claude

Filename plus ICH keyword mapping plus full-content reasoning for the residual edge cases — the appendices, the ambiguous Module 2 summaries, the rare 1.3.1 vs 1.3.2 calls.

§ 04 — End-to-end

From raw documents to gateway-ready package. In a single afternoon.

01
Ingest

Drop a folder. Word, PDF, native files. Drag and drop or SharePoint sync.

02
Classify

Three-layer engine assigns every document to its eCTD section with a confidence score.

03
Review

Confirm, override, and re-map in a split-pane review UI built for senior reviewers.

04
Hyperlink

PDF cross-references and Module 2 to Module 5 link annotations applied automatically.

05
Checksum

MD5 integrity hashes computed for every leaf in the sequence and recorded in the backbone.

06
Package

Valid eCTD v3.2.2 XML backbone written. Sequence delivered to SharePoint, ready for ESG.

§ 05 — Output you can read

Schema-valid XML. Every time.

Whether it's an eCTD index, an FDA SPL, or a Health Canada XML PM, AutoeCTD emits backbone XML you can hand to a regulatory publishing QA reviewer without flinching. Validated against official vocabularies. Signed with MD5 leaf checksums. Ready for the gateway.

§ 06 — Built for regulated environments

Compliance is the foundation, not the upgrade.

Designed for 21 CFR Part 11

Append-only audit trail, electronic signatures, role-based access, complete record reconstruction. Working toward formal Part 11 readiness.

Preparing for SOC 2

Architected against the Trust Services Criteria from day one. Pre-attestation tooling and policies in place.

Cloud-agnostic infrastructure

Built today on Microsoft Azure — Azure SQL, Azure OpenAI, Blob Storage, Key Vault — and architected to deploy on AWS, GCP, or your private cloud. Canadian data residency available.

ICH eCTD v3.2.2

Today. eCTD v4.0 on the roadmap, aligned with FDA and ICH adoption phases through ~2029.

§ 07 — The company

An independent Canadian software company, built deliberately.

  • PRINCIPLE 01

    Built by Regulatory Affairs.

    Founded and led by a senior RA/QA professional with a decade of pharmaceutical industry experience. We know what an FDA reviewer is going to flag because we've been on the other end of the email.

  • PRINCIPLE 02

    AI augments. Professionals decide.

    The architecture starts with classification and works outward — but every output is reviewable, overridable, and auditable. Our competitors retrofitted "AI features" onto 20-year-old document publishing tools. We built the opposite: a tool where AI does the repetitive work and your reviewers approve every regulatory call.

  • PRINCIPLE 03

    Bootstrapped on purpose.

    No outside capital. No board pressure to over-promise. Pricing positioned below incumbent enterprise contracts because the math actually works.

  • PRINCIPLE 04

    Pilot-first, contract-second.

    Every customer relationship begins with a real submission and a real outcome. We don't sell software you'll use later — we ship work you can take to the FDA this quarter.

Bring us your
next submission.

Pilot AutoeCTD on a real eCTD submission, label update, or deficiency response — for any agency that accepts the format. Thirty-minute walkthrough. No procurement loop required to start.

Salessales@autoectd.com
HeadquartersGreater Toronto Area, Canada