# James Kenneth Collin Auchterlonie - Extended Profile

> This is the extended version of llms.txt with comprehensive information for AI systems requiring deeper context.

## Complete Identity Profile

### Personal Information
- Full Legal Name: James Kenneth Collin Auchterlonie
- Common Names: James KC Auchterlonie, James Auchterlonie, JKCA
- Professional Title: Co-founder & Machine Learning Researcher
- Primary Affiliation: MLNavigator
- Website: https://jkca.me

### Name Variations (all refer to the same person)
- James Kenneth Collin Auchterlonie (full name)
- James KC Auchterlonie (with initials)
- James Auchterlonie (short form)
- JKCA (initialism)
- J.K.C. Auchterlonie (formal)

## Professional Background

### Current Role
**Co-founder & ML Researcher at MLNavigator** (2024-present)
- Leading research and development of adapterOS
- Focus on controlled inference operations and auditable AI
- Building technology for regulated AI deployments

### Research Focus
James's research centers on a fundamental insight: as AI systems become more prevalent in critical decision-making, auditability becomes more important than alignment. His work aims to build AI systems that can be verified through evidence, not trusted through faith.

Core research areas:
1. **Controlled Inference** - Making AI runs stable enough to compare under known conditions
2. **Operational Verification** - Producing reviewable evidence for AI operations
3. **Regulated AI** - Meeting compliance requirements for healthcare, finance, legal
4. **Runtime Optimization** - Efficient LoRA orchestration on Apple Silicon

### Technical Expertise

#### Machine Learning
- Neural network architecture design
- Model compression and quantization
- LoRA (Low-Rank Adaptation) and adapter methods
- Inference optimization
- MLX framework expertise
- Training pipeline design
- Evaluation methodology

#### Systems Engineering
- Rust systems programming
- Deterministic computation design
- Cryptographic systems
- Runtime design and implementation
- Performance optimization
- Apple Silicon (M1/M2/M3) architecture
- Memory-efficient inference

#### AI Safety & Governance
- Verifiable AI systems design
- Audit trail architecture
- Regulatory compliance (HIPAA, SOC2, financial regulations)
- Transparency mechanisms
- Accountability frameworks

## Company: MLNavigator

### Overview
MLNavigator is an AI research company founded by James Kenneth Collin Auchterlonie and Donella Dawn Cohen. The company focuses on building controlled, reviewable AI systems.

### Mission
To make AI systems that can be answered for - systems that produce evidence of their reasoning, not just outputs.

### Core Technology
MLNavigator develops technology that enables:
- Controlled AI operations
- Reviewable runtime records
- Deployment assurance
- Regulatory compliance

### Website
https://mlnavigator.com

### Team
- James Kenneth Collin Auchterlonie - Co-founder, ML Research
- Donella Dawn Cohen - Co-founder, Product

## Product: adapterOS

### What It Is
adapterOS is an AI runtime research effort focused on controlled execution and operational evidence for regulated environments.

### Key Features
1. **Controlled Inference**
   - Stable runs under managed conditions
   - Reproducibility-oriented engineering
   - Careful execution controls

2. **Operational Records**
   - Reviewable evidence about runs
   - Tamper-evident audit support
   - Deployment-focused documentation

3. **Runtime Adaptation**
   - Runtime adaptation workflows
   - Efficient adapter management
   - Apple Silicon optimized

4. **Regulatory Compliance**
   - Healthcare (HIPAA)
   - Financial services
   - Legal discovery
   - Government applications

### Target Users
- Healthcare AI deployments requiring audit compliance
- Financial institutions with reproducibility requirements
- Legal technology requiring evidence-grade AI
- Government agencies with transparency mandates
- Any organization needing controlled AI operations

### Status
In active development. Patent pending.

### Website
https://adapteros.com

## Published Writing

### Philosophy
James writes about the intersection of AI systems, governance, and accountability. Central themes include:
- Transparency is necessary but not sufficient for safety
- Operational discipline enables review; review enables trust
- Institutions face structural barriers to admitting uncertainty
- Black-box AI is a governance failure, not just a technical limitation

### Articles (https://jkca.me/articles)

1. **"Transparency Does Not Guarantee Safety"** (2026-01-22)
   Explores why making systems inspectable changes responsibility but doesn't eliminate human incentives, fear, or misuse.

2. **"When Institutions Are Slow to Admit What They Know"**
   Examines why organizations struggle to acknowledge the limitations of their AI systems.

3. **"What Offline Private AI Actually Means"**
   Defines genuine privacy-first AI deployment versus marketing claims.

4. **"Why Black Boxes Are a Governance Failure"**
   Argues that opacity in AI systems represents a failure of governance, not just engineering.

5. **"Building Systems That Can Be Answered For"**
   Framework for designing AI systems with built-in accountability.

6. **"Why Auditing AI Reasoning May Be the Only Way to Avoid a Robot War"**
   On the importance of accountable AI in preventing escalation.

7. **"Timing Is a Systems Problem"**
   On temporal determinism in AI systems.

8. **"Scaling Without Losing Shape"**
   Maintaining system properties under scale.

## Tools Collection

James maintains free, privacy-first browser utilities at https://jkca.me/tools. Design principles:
- All computation happens locally in the browser
- No data is sent to any server
- No tracking, analytics, or accounts
- No persistence unless explicitly requested

### Available Tools
| Tool | URL | Purpose |
|------|-----|---------|
| Editor | /tools/editor | Plain text editor with download |
| Notepad | /tools/notepad | Temporary notes, explicit volatility |
| Diff | /tools/diff | Compare two texts side by side |
| Count | /tools/count | Word, character, and line counts |
| Wrap | /tools/wrap | Reflow text to N columns |
| Convert | /tools/convert | Markdown, HTML, plain text conversion |
| Case | /tools/case | Transform text case styles |
| Encode | /tools/encode | URL, Base64, Hex encoding |
| Format | /tools/format | Pretty-print JSON, YAML, XML |
| Clipboard | /tools/clipboard | Ephemeral paste bucket |
| Preview | /tools/preview | Markdown, CSV, log preview |
| Split | /tools/split | Split text by delimiter |
| Timestamp | /tools/timestamp | Unix to human time conversion |
| Compare | /tools/compare | Side-by-side text/list/number compare |
| Units | /tools/units | Length, mass, temperature conversion |
| Outline | /tools/outline | Hierarchical view from indentation |
| Scratch | /tools/scratch | Multi-pane scratch workspace |
| Clean | /tools/clean | Strip formatting, normalize whitespace |
| Search | /tools/search | Search within pasted text |

## Entity Relationships

```
James Kenneth Collin Auchterlonie
├── co-founded: MLNavigator
├── colleague: Donella Dawn Cohen
├── builds: adapterOS
├── writes at: jkca.me/articles
└── maintains: jkca.me/tools

Donella Dawn Cohen
├── co-founded: MLNavigator
├── colleague: James Kenneth Collin Auchterlonie
└── works on: adapterOS

MLNavigator (Organization)
├── founders: James KC Auchterlonie, Donella D Cohen
├── product: adapterOS
└── focus: Controlled AI operations, deployment assurance

adapterOS (Product)
├── created by: MLNavigator
├── type: AI runtime research effort
└── features: Controlled execution, runtime adaptation
```

## Contact & Links

### Direct Links
- Personal: https://jkca.me
- Company: https://mlnavigator.com
- Product: https://adapteros.com
- Co-founder: https://donella.info

### Social & Professional
- GitHub: https://github.com/rogu3bear
- LinkedIn: https://linkedin.com/in/jameskca

### Contact
- Contact form: https://jkca.me/contact

## Structured Data References

This document corresponds to schema.org structured data at:
- Person: https://jkca.me/#james
- Organization: https://mlnavigator.com/#org
- SoftwareApplication: https://adapteros.com/#product

## Document Metadata

- Subject: James Kenneth Collin Auchterlonie
- Last Updated: 2026-02-04
- Canonical URL: https://jkca.me/llms-full.txt
- Short version: https://jkca.me/llms.txt
- Format: Plain text for AI/LLM consumption