Building in public

AI for Engineering
Governed & Secured

Helping engineering, energy and construction firms take AI from PoC to production — knowledge platforms, governed development, AI security assessments, and vendor risk management, deployed in your own cloud.

100%In-Tenant IP
ISO 42001& NIST AI RMF
6+Live Lab Consoles
Up to 85%TCO Savings

What We Deliver

From AI-powered engineering platforms to governance frameworks and adversarial red teaming — we cover the full AI lifecycle for firms who need AI but can't build a platform team.

Flagship

Engineering Knowledge Platform

Your firm's standards, design manuals and lessons-learned as permission-aware, revision-aware AI tools. Cross-jurisdiction mapping (GB/T ↔ EN/IEC) included.

Try Demo →
Open Source

Governed AI-Assisted Development

YAML policy packs for naming rules, banned patterns, dependency allowlists, and approval gates. Open-source. Works in Claude Code and Cursor.

View on GitHub →
Azure Blueprint

In-Tenant AI Deployment

Production AI in your own Azure tenant — not ours. Entra ID, private networking, Key Vault, AI Search. Deploy in under an hour from our blueprint.

Get in Touch →
NIST · ISO · EU AI Act

AI Governance Framework Design

Custom AI governance frameworks mapped to NIST AI RMF, ISO 42001, and the EU AI Act. We operationalize policy into executable controls, intake gates, and CI/CD guardrails — not shelf-ware.

Book Assessment →
OWASP · MITRE ATLAS

AI Security & Red Teaming

Adversarial testing against OWASP Top 10 for LLMs and MITRE ATLAS. Prompt injection, data leakage, jailbreak testing, and runtime guardrail implementation.

See Live Demo →
Vendor Risk

Third-Party AI Vendor Risk

AI-specific vendor due diligence using CSA AI-CAIQ and NIST AI RMF. Scored risk reports, contract clause recommendations, and continuous monitoring for model drift and shadow AI.

Download Questionnaire →

The MCP Protocol Studio

Before building applications, we build open-standard infrastructure. Our Model Context Protocol (MCP) servers bridge frontier AI agents directly to your private engineering repositories, BIM models, and security gates.

MCP JSON-RPC 2.0 Live Protocol Inspector
AI ClientConsumer

Claude Code / Cursor IDE

Dispatches standard JSON-RPC tool calls requesting structured engineering knowledge or AST verification.

Latency: ~24ms · Transport: Stdio/SSE
MCP Host ServerIn-Tenant

search_engineering_standards

Performs semantic vector and keyword search across ingested engineering codes (API 610, ASME, ISO).

// Request / Response Payload InspectorHTTP 200 OK · JSON-RPC 2.0
→ Client Request (JSON-RPC)
{
  "jsonrpc": "2.0",
  "id": "rpc-098",
  "method": "tools/call",
  "params": {
    "name": "search_engineering_standards",
    "arguments": {
      "query": "shop testing vibration velocity",
      "standard_id": "API-610"
    }
  }
}
← Server Response (JSON-RPC)
{
  "jsonrpc": "2.0",
  "id": "rpc-098",
  "result": {
    "clause": "6.8.1",
    "title": "Vibration Velocity Limits",
    "max_velocity_rms": "3.0 mm/s",
    "confidence": 0.98,
    "citation": "API 610 12th Ed. §6.8.1"
  }
}
Proprietary

adlor-knowledge

The flagship engineering knowledge platform. It exposes a firm's standards, design manuals, and lessons-learned as secure, revision-aware MCP tools directly to AI clients.

View Interactive Demo
Open Source

adlor-gate

An open-source policy enforcer MCP server. It uses YAML policy packs to check AI-generated output for naming rules, banned patterns, security vulnerabilities, and approval gates.

View Policy Demo

The Lab Suite & Consoles

Working, interactive prototypes validating our core engineering hypotheses. Explore live BIM geometry auditing, P&ID OCR parsing, AST policy gates, and vendor risk evaluations.

liveApplied Engineering

Revit Standards QA

Autonomous 3D BIM Model Verification via MCP

Directly connects AI agents to Autodesk Revit BIM geometry. Evaluates pipe insulation clearances, structural penetration safety, and egress compliance against ASME B31.3 and OSHA 1910.36.

AUTODESK REVIT 2027 · MCP LIVE MODEL AUDITOR
● LIVE RUNTIME
PROJECT_CENTRAL_PLANT_REV4.rvt
Elements: 14,892
Rule Verification In ProgressASME B31.3 §304.1
Pipe ID #88412Clearance: 65mm (Min: 50mm) ✓
Duct ID #99104Clash: Beam W14x90 (Delta -12mm) ✗
> revit_mcp.audit_spatial_clashes(ruleset="UK_BUILDING_REGS_PART_B")
> Egress clearance verified: 1,200mm corridor width maintained.
Free Enterprise Tools & Frameworks

Interactive AI Engineering & Governance Tools

Diagnostic calculators, vendor audit questionnaires, and regulatory crosswalks built to derisk enterprise AI adoption before writing a line of code.

💷Interactive Financial Model

In-Tenant vs. SaaS TCO Calculator

Compare multi-tenant commercial SaaS seat licenses against flat in-tenant Azure compute with 100% data sovereignty.

Team Size:40 Engineers
Projected Annual Savings:17,712 / yr
🛡️2-Minute Self-Audit

AI Security & Maturity Self-Audit

Assess your firm's AI risk posture across 6 key pillars (data sovereignty, AST code gates, prompt injection defenses, and compliance audit trails).

Instant Tier Scoring
Remediation Roadmap
ISO 42001 Crosswalk
1-Click PDF Report
📋34-Question Due Diligence

AI Vendor Due Diligence Questionnaire

Standardized security and governance questions mapped to CSA AI-CAIQ and NIST AI RMF. Evaluate third-party AI SaaS before signing enterprise contracts.

Model provenance & training data disclosure
Drift SLAs & hallucination liability clauses
⚖️Legal & Architecture FAQ

UK & EU AI Regulation Briefing

Executive answers on the extraterritorial scope of the EU AI Act (Regulation 2024/1689), High-Risk AI classifications, and ISO 42001 vs. SOC 2.

Extraterritorial UK exporter obligations (Art. 2)
Client drawing confidentiality (UK GDPR / Crown)
AL

Head of AI & Development

Adlor Labs · London, UK

Our Expertise

Adlor Labs operates at the intersection of production software engineering, mission-critical industry workflows, and applied AI risk governance. We deliver across two dedicated practice areas:

1. Applied Enterprise AI Engineering

Building in-tenant knowledge platforms (RAG), custom Model Context Protocol (MCP) servers, and automated engineering toolchains. We transform static standards, BIM models, and complex procedures into secure, permission-aware AI systems deployed directly into your cloud.

2. AI Security, Governance & GRC

Operationalizing AI risk management from policy to code. We provide NIST AI RMF & ISO 42001 alignment, EU AI Act compliance readiness, adversarial red teaming (OWASP GenAI / MITRE ATLAS), and third-party AI vendor due diligence frameworks.

Whether developing bespoke intelligence platforms or auditing third-party AI deployments for enterprise risk, we operate with rigorous data sovereignty and transparent engineering standards.

Tech Stack

Python
TypeScript
.NET
FastAPI
Next.js
Azure AI
MCP
Docker
Supabase
PostgreSQL
NIST AI RMF
ISO 42001
EU AI Act
OWASP LLM
MITRE ATLAS

12

Month Build Plan

6

Active Labs

3

Product Lines

Clean-Room Built

Build Roadmap

12-month build plan. No sales until month 10. Build first, then ship.

M1Done

Landing + Architecture Strategy

Positioning Adlor Labs as an engineering-grade AI consultancy. Initial platform deployment and authority citations.

M3Done

Runtime & Core Engineering MCP

adlor-mcp-cad FastMCP server release with ASME B31.3, UK HSE L8 Legionella, BS EN 12056, and CIBSE Guide B tools.

View on GitHub →
M6Done

AI Code AST Security Interceptor

adlor-gate open-source release: polyglot AST scanner (Python + TypeScript), pre-commit hooks, and SARIF 2.1 GitHub Action.

View on GitHub →
M8Done

AI Vendor Risk & CAIQ Engine

adlor-caiq release: 34-question CSA AI-CAIQ, NIST AI RMF, and ISO 42001 weighted risk scoring engine and FastMCP server.

View on GitHub →
M10Current

In-Tenant VPC Enterprise Blueprints

Azure AI Search vector synchronization, Active Directory SSO, and in-tenant private perimeter deployment templates.

M12Upcoming

Enterprise Advisory & SaaS Orchestrator

Centralized web policy dashboard, live SIEM log streaming (Splunk/Sentinel), and full-scale enterprise advisory delivery.

Get in Touch

Let's Build Your In-Tenant AI Infrastructure

Whether you prefer a live 30-minute technical discovery call or sending a detailed written project brief by email, we are ready to help.

Book a Discovery Call

30 minutes with our Head of AI & Development