AIMS
AI-Implemented & Maintained Software
By Maxime Leboeuf & Bruno Potvin, Co-founders — Somtech | February 2026
60-80%
Cost reduction
4-8x
Faster
100%
Client ownership
24/7
AI Maintenance
Table of contents
Executive Summary
Custom software has historically been reserved for organizations that can invest hundreds of thousands of dollars and wait months for development. AIMS — AI-Implemented & Maintained Software — changes this reality. In this model, artificial intelligence performs all technical work (architecture, development, testing, deployment, maintenance), while humans focus on strategic oversight and business validation.
This white paper defines AIMS as an industry standard: its principles, methodology, architecture, and quality criteria. The goal is to provide a reproducible reference framework for any organization wishing to adopt this model.
Note: The AIMS acronym is intentionally in English to reflect the international scope of the standard. This document is written in French, as Somtech is a Quebec-based company.
Context: Why AIMS Now
2.1 The Limits of the Traditional Model
Custom software development relies on a model that is decades old: a team of developers, architects, testers, and project managers work together to translate business needs into functional code. This model has proven its worth, but its structural limitations are becoming increasingly problematic.
- Cost :A custom project typically costs between $50,000 and $500,000+ for an SMB. At $100–200/hour per developer, budgets escalate quickly.
- Timelines :Development cycles stretch over months or even years. Delays accumulate and budgets spiral.
- Misalignment :The client describes their reality on day one, but receives software months later, when their processes and market have already evolved. The delivered product no longer matches the actual need.
- Maintenance :After delivery, maintenance represents 60–80% of the total cost of ownership. Every update requires costly human intervention.
- Dependency :System knowledge resides in developers' heads, not in living documentation. Technology lock-in is the norm.
- Talent shortage :The industry faces a global shortage of qualified developers. In Quebec alone, thousands of IT positions remain unfilled, delaying the operational evolution of businesses.
2.2 The Technological Convergence
Several advances converge to make AIMS possible today:
- Advanced Language Models (LLMs): Large language models now generate production-quality code, understand complex architectures, and reason about technical problems.
- Headless Autonomous AI Agents: Agents operate directly from the command line — they analyze a codebase, plan modifications, write code, run tests, and deploy autonomously.
- Mature Cloud Infrastructure: Continuous deployment, infrastructure as code, and managed services enable full automation of the delivery pipeline.
- Standardized Connection Protocols (MCP): The Model Context Protocol allows AI agents to communicate directly with cloud services, databases, and deployment tools, eliminating manual integrations.
- Automated DevOps Pipelines: CI/CD pipelines and hosting platforms offer programmatic interfaces that AI agents can orchestrate natively.
2.3 The AIMS Moment
AIMS is not a futuristic prediction. It is the formalization of a model already in practice. Somtech, founded in Quebec by Maxime Leboeuf and Bruno Potvin, applies this approach in production with real clients. This white paper aims to codify the model into an open standard, adoptable by the entire industry.
Definition of AIMS
3.1 The Acronym Decoded
AIMS stands for AI-Implemented & Maintained Software. Each component carries a precise meaning:
Artificial intelligence is the primary execution agent. These are not assistance tools — they are the direct executors of technical tasks.
Implementation covers the entire construction cycle: technical analysis, architecture, coding, testing, infrastructure configuration, deployment. Everything is executed by AI.
Continuous maintenance is handled by AI: bug fixes, security updates, feature enhancements, performance optimizations. The software evolves continuously without human technical intervention.
The deliverable is a functional, deployed software that the client fully owns. Not a prototype — a product in production.
3.2 Formal Definition
AIMS is a software development and maintenance model in which artificial intelligence assumes all technical execution — from architecture to deployment and ongoing maintenance — under the strategic supervision of a human architect who validates business decisions and ensures alignment with the client's objectives.
3.3 What AIMS Is and What It Is Not
| AIMS is | AIMS is not |
|---|---|
| A model where AI executes all technical work | A developer assistance tool (copilot) |
| A system supervised by a strategic human architect | A total replacement of human expertise |
| An end-to-end model (development + maintenance) | A one-time code generator (no-code/low-code) |
| A reproducible standard for any qualified firm | A proprietary closed technology |
| An engagement where the client owns 100% of the code | A SaaS model where the code stays with the vendor |
Continue reading
Enter your email to receive an access link to the remaining 11 chapters and the full PDF.
« The future of software development will not be defined by the amount of code a developer can write, but by the quality of the vision an architect can direct. »
— Maxime Leboeuf, Co-founder, Somtech
Ready to discover how AIMS can evolve your operations? Book a free consultation with one of our architects.