# keli-ia-generative-code-100-percent-approche-2025-11-05

## Veille

Functional approach to generative AI in software development, 100% generated code, LLM onboarding, atomic tasks, spec-driven, continuous capitalization - Soufiane Keli - OCTO Technology - LinkedIn

## Titre Article

Approche fonctionnelle pour l'IA générative en développement : 100% de code généré

## Date

2025-11-05

## URL

https://www.linkedin.com/feed/update/urn:li:activity:7391722941790527489/

## Keywords

generative AI, code generation, software development, atomic tasks, spec-driven development, LLM onboarding, rapid iteration, continuous capitalization, brownfield, code standards, structured prompts, DoD, OCTO Technology, velocity, code quality, junior developer, global context, user stories, LLM

## Authors

Soufiane Keli (OCTO Technology)

## Ton

**Profile:** Conversational-Professional | First-person practitioner | Educational-Promotional | Intermediate-Technical

Keli, an OCTO consultant, shares a concrete methodology with an experienced practitioner's tone. A four-step structured approach ("AI Onboarding", "atomic tasks", "spec-driven", "continuous capitalization") reveals a teaching stance combining good practices and brownfield field experience. The language alternates between methodological prescriptions and pragmatic admissions ("Don't ask the AI to do everything at once"). Typical of tech consultants on LinkedIn sharing field expertise with soft promotion of their company.

## Pense-betes

- **AI generates nearly 100% of code** through a structured approach
- **Step 0: AI Onboarding** - treat the LLM like a "junior every day", restate global context
- **Step 1: Exploration & Planning** - break user stories down into precise atomic tasks
- **Use the LLM for the breakdown** - reach the right level of detail
- **Step 2: Iterative spec-driven development** - structured prompts + code examples + standards + DoD checklist
- **Dynamic adjustment** - modify the prompt/context if unsatisfied
- **Step 3: Continuous capitalization** - enrich the good-practices base after each cycle
- **Key principle**: atomic tasks + clear standards + rapid iteration = velocity AND quality
- **Don't ask the AI to do everything at once** - incremental approach
- **Brownfield demonstration** - Loïc Lefloch and Simon Belbeoch (OCTO)
- **Pragmatic approach** combining existing good practices

## RésuméDe400mots

Soufiane Keli, a consultant at OCTO Technology, proposes a methodical approach that allows generative AI to produce nearly 100% of the code in a real software project. Far from magical, this strategy rigorously combines proven practices within a process structured into four steps.

**Step 0: Daily LLM Onboarding**

Unlike a human developer who retains context across sessions, the LLM must be "rehired" every day. This crucial step consists of systematically restating the project's global context, its business and architectural objectives. Treating the model as a "junior who starts over every morning" enforces a beneficial documentation discipline: explicit context becomes an asset shared by the whole team, not merely tacit knowledge held by a few.

**Step 1: Exploration and Atomic Planning**

Before any generation, each user story undergoes a meticulous breakdown into atomic tasks with precise descriptions. Paradoxically, this planning itself uses an LLM to identify the optimal level of granularity. This step transforms vague functional objectives into actionable technical specifications, drastically reducing the ambiguity that models handle poorly.

**Step 2: Iterative Spec-Driven Development**

Development proper relies on highly structured prompts comprising four elements: a detailed technical specification, illustrative code examples, project standards and conventions, and an explicit Definition of Done (DoD) checklist. If the result is unsatisfactory, the approach prescribes a methodical adjustment of either the prompt or the context, avoiding random iterations. This rigor turns interaction with the LLM from an improvised conversation into a reproducible engineering process.

**Step 3: Continuous Capitalization**

After each cycle, the lessons learned progressively enrich an organizational knowledge base: successful patterns, effective prompt formulations, identified pitfalls, documented examples. This continuous improvement loop turns individual experience into collective intellectual capital, accelerating future projects.

**Fundamental Principle and Field Validation**

The guiding principle explicitly rejects monolithic generation: "Don't ask the AI to do everything at once". Instead, atomic tasks + clear standards + rapid iteration simultaneously produce velocity AND quality—objectives traditionally seen as antagonistic.

Crucially, this approach was demonstrated in a brownfield environment by Loïc Lefloch and Simon Belbeoch at OCTO Technology, proving its applicability beyond idealized greenfield projects. The brownfield context—with legacy code, existing architectural constraints, technical debt—represents the real working ground for the majority of developers.

**Pragmatic Positioning**

Keli explicitly positions this method as a pragmatic combination of publicly available good practices, not as a radical innovation. This strategic modesty reinforces its credibility: the approach does not require a cultural revolution, only a disciplined orchestration of known techniques adapted to the generative-AI context.

## GrapheDeConnaissance

- Soufiane Keli —travaille_chez→ OCTO Technology (ORGANISATION, 0.98)
- Soufiane Keli —publie→ approche spec-driven IA (METHODOLOGIE, 0.98)
- approche spec-driven IA —permet→ 100% code généré par IA (CONCEPT, 0.92)
- approche spec-driven IA —utilise→ onboarding quotidien LLM (CONCEPT, 0.95)
- approche spec-driven IA —utilise→ tâches atomiques (CONCEPT, 0.97)
- approche spec-driven IA —utilise→ capitalisation continue (CONCEPT, 0.93)
- Loïc Lefloch —soutient→ approche spec-driven IA (METHODOLOGIE, 0.95)
- Simon Belbeoch —soutient→ approche spec-driven IA (METHODOLOGIE, 0.95)
- approche spec-driven IA —s_applique_à→ environnement brownfield (CONCEPT, 0.93)
- tâches atomiques —améliore→ vélocité et qualité (CONCEPT, 0.9)
- Definition of Done —fait_partie_de→ prompt structuré (CONCEPT, 0.92)
- Yves Caseau —soutient→ approche spec-driven IA (METHODOLOGIE, 0.85)
- Yves Caseau —travaille_chez→ Michelin (ORGANISATION, 0.97)

---
Canonical: https://www.thekb.eu/en/fiches/keli-ia-generative-code-100-percent-approche-2025-11-05/