Statistical programming that grows into submission leadership.

ME-TA — Martin Eeg and Toni Awwad — deliver TFLs, ADaM, and SDTM with the depth to direct your submission when it matters. Built from firsthand experience inside statistical departments at Novo Nordisk and Novartis, backed by a validated platform and AI-augmented production.

Worked within Novo Nordisk & Novartis | 40+ Projects | 100,000+ Outputs | 21 CFR Part 11 Compliant

How We Work

We start as your statistical programmers. When complications arise, we bring the depth to direct the submission.

Entry Point

Statistical Programming

TFLs, ADaM, SDTM — the hands-on work

  • Martin produces TFLs and ADaM, Toni produces SDTM, TFLs, and ADaM
  • AI-augmented production on Metawebservice or in your environment
  • Portable SAS 9.4 programs with validation package
  • Typically 50-65% below traditional CRO pricing
Trust Earned

Submission Leadership

When the submission gets complicated

  • Statistical strategy and regulatory argumentation
  • Integrated analyses (ISS/ISE) for regulatory submissions
  • SAPs, CSR statistics sections, DSMB support
  • The depth that gets products approved at FDA
Full Service

Complete Biometrics

For companies without their own SCE

  • Programming + strategy + platform in one engagement
  • Metawebservice provides validated storage, compute, and audit trail
  • Cross-study data management and pooling
  • Real-time visibility into all deliverables

What Becomes Possible

Autonomous AI agents can now write SAS from specifications, execute programs, read logs, and iterate until clean. But they need infrastructure that supports them. Most pharma SCEs were built for humans working through remote desktops and GUIs — environments where agents cannot operate. Metawebservice is API-first: designed so AI agents and statisticians work side by side on the same platform.

Safety Oversight

DSMB Members Arrive Prepared

Safety officers, clinical operations, DSMB chairs

Board members meet every three to six months. Between meetings, they return to their full-time roles and lose the thread. When the next data package arrives, they spend hours re-reading the last one to remember what was flagged, what was decided, and what was being watched. The statistician preparing the briefing spends days manually compiling what changed since last time.

  • AI maintains the full meeting history as living study context — every prior data package, every decision, every action item — and generates an executive summary so members pick up exactly where they left off
  • AI classifies the most critical findings in each new data cut: new safety signals, trending adverse events, enrolment shifts — ranked so members focus on what matters instead of scanning 100 tables
  • Every member's data access is logged and auditable — who viewed what, when — with all materials version-controlled per data cut
  • Between meetings, the study context keeps accumulating — protocol amendments, new sites, data queries — so nothing falls through the cracks when the board reconvenes
Knowledge Continuity

New Team Members Productive in Days, Not Weeks

Device companies with multiple concurrent studies and high turnover

Someone leaves and takes the study knowledge with them. The new hire inherits a shared folder with hundreds of files and no narrative — no record of what was done, why, or what still needs to happen. The first weeks are archaeology: reading old programs, guessing at intent, occasionally re-doing work that already exists somewhere.

  • AI builds a living knowledge base per study — updated automatically with each new analysis, decision, and deliverable — so the study's history is always current and always readable
  • When a new team member joins, AI generates a study orientation brief: current status, recent activity, pending deliverables, key decisions, and pointers to the relevant programs and outputs
  • When a new analysis is requested, AI surfaces related prior work — "this was done before, here are the results and the reasoning" — preventing duplicate effort across personnel changes
  • All context lives at study level, not in someone's inbox or on their laptop — the company's analytical knowledge survives any departure
First Submission

500 TFLs Without Building a Department

Virtual biotech preparing their first NDA or MAA

The trial succeeded, the deadline is set, and the team is three people who have never built a submission package. A traditional CRO quotes 12–18 months with milestone-only visibility. The CEO asks "are we on track?" every week and nobody can answer until the next quarterly report.

  • AI holds the full study context — protocol, SAP, shell specifications, data structure — and generates SAS programs through an autonomous loop: write, execute, read logs, iterate until clean
  • As deliverables are produced, AI tracks submission readiness in real time: which TFLs are complete, which are in QC, which are blocked and why — visible to the whole team, always current
  • Structural QC by AI checks every output against shell specifications — headers, footnotes, precision, sort order — the systematic checking that otherwise consumes a third of a programmer's time
  • When specifications change or FDA feedback requires new analyses, the AI already has the full context — it knows what exists, what is affected, and what needs to be created

21-Policy Quality Management System

ME-TA operates a comprehensive QMS under ICH E6/E9, 21 CFR Part 11, EU Annex 11, and GDPR. All policies are version-controlled in audit-ready format.

4-Hour Breach Notification

ME-TA commits to notifying affected parties within 4 hours of a confirmed breach — far exceeding the 72-hour GDPR minimum requirement.

Infrastructure as Code

All infrastructure changes are deployed through version-controlled code with two-person approval gates before reaching production.

SANS-Based Incident Response

Incident response procedures follow the SANS framework with annual testing to ensure readiness and effectiveness.

Audit-Ready at All Times

Version-controlled policies, complete audit trails, and documented change control. We welcome vendor audits — request an audit slot at any time.

How We Use AI

AI makes ME-TA faster, not less secure. Your data stays where you expect it. A dedicated AI Usage Policy (P22) governs all AI activities.

What AI Handles

  • Code generation from specifications (SAP to SAS programs)
  • Structural QC: output format compliance, cross-referencing specifications
  • Documentation drafting: define.xml, reviewers' guides, annotated CRFs

What AI Never Touches

  • Patient-level data is never sent to external AI services
  • Randomization codes and unblinding information
  • Personally identifiable information (PII)

Enforced by Architecture, Not Just Policy

The Metawebservice desktop client runs in a dedicated AI mode where the sync mechanism only downloads SAS programs and reviewed context files. Patient data, datasets, and outputs are never synced to the AI workstation. AI tools on that machine have full local access — because there is nothing sensitive to access. The data boundary is enforced by what the server sends, not by what the AI is told to ignore.

API-First: Built for Agents

Most SCEs are locked behind GUIs — remote desktops, SAS Enterprise Guide, Citrix sessions. Metawebservice is API-based: 159 REST endpoints covering file operations, SAS job execution, dependency tracking, and audit logging. AI agents can programmatically upload programs, trigger execution, read logs, and download outputs — enabling end-to-end automated workflows that are simply not possible on traditional compute environments.

Human Oversight

All AI-generated outputs go through the same two-level QC process as human-generated work. Every deliverable is reviewed by a senior statistician.

Security Architecture

Enterprise-grade security built on AWS, designed to satisfy pharma IT and regulatory requirements.

EU Data Residency

AWS eu-west-1 (Ireland). All data remains within the EEA at all times.

Encryption

AES-256 encryption at rest. TLS 1.2+ for all data in transit.

Authentication

AWS Cognito with Email OTP or SSO integration. Frictionless yet secure.

Access Control

Role-based access with customer-controlled study permissions and project isolation.

Audit Logging

AWS CloudTrail plus application-level audit logging. Immutable, tamper-evident records.

Cloud Infrastructure

Built entirely on AWS managed services. Infrastructure deployed as code with full change control.

Metawebservice Platform

A .NET Core API with 159 endpoints purpose-built for clinical biometrics. A Windows client that integrates directly with your local SAS installation. Automated sync, file tracking, execution scheduling, and a complete audit trail — all in one platform.

API & Backend

  • 61 SAS integration endpoints (job execution, scheduling, dependency tracking)
  • 17 file operation endpoints (upload, download, versioning)
  • Study management with role-based permissions
  • Context file management and audit logging
  • Study invitations and customer administration
  • OpenAPI 3.0 documented (Swagger UI at API root)

Desktop Client

  • Windows application installed on each data processor
  • Integrates with local SAS 9.4 installation
  • Automated sync: local changes detected and uploaded
  • File tracking with MD5 hash verification
  • SAS job execution, monitoring, and log analysis
  • Deployed via Microsoft Intune (MSIX)

Discuss Your Project

Deliverables

AI-augmented production, expert-reviewed. Portable SAS 9.4 programs + outputs + validation package. Typically 50-65% below traditional CRO pricing.

Clinical Listings

Patient data listings produced from shells. Highly automatable — AI generates the code, expert reviews the output.

SDTM + ADaM

Raw data to submission-ready datasets. SDTM mapping, ADaM derivations, define.xml, and validation package.

Integrated Analyses (ISS/ISE)

Pool across studies: SDTM harmonization, pooled ADaM, safety TFLs. Cross-study work on Metawebservice with full traceability.

Request a Quote

Track Record

40+
Projects Delivered
5,000+
Files Uploaded
100,000+
Outputs Created
20+
Years in Pharma
21
Quality Policies
4 hrs
Breach Notification SLA

GCP CVs available on request.

Get in Touch

Contact

contact@me-ta.dk

Address

Agern Alle 5A

DK-2970 Horsholm, Denmark

Company

ME-TA ApS

CVR: 37265705

How can we help?

QA / IT Security

Request the qualification package or schedule a vendor audit.

Request Qualification Package

Biometrics

See the platform in action with a live walkthrough.

Request a Platform Demo

General Inquiries

Questions about services, pricing, or anything else.

Request More Information