# rafal-wenvision-ia-generative-produit-techno-pas-projet-2024-02-23

## Veille

Op-ed by **Olivier Rafal** (Consulting Director Strategy at **WeNvision**) published on **February 23, 2024** on **CIO-Online** (*Tribune* section), advancing a thesis still counter-intuitive at the time: **generative AI is more a matter of technology product than an AI/data science project**. **Argument 1 — data science is not the core issue**: building a *foundation model* from scratch requires *« several months, millions of euros, and access to enormous quantities of data »* — reserved for players with specific, monetizable datasets (e.g. **Bloomberg** and its **BloombergGPT** for finance). For nearly all companies, the right reflex is therefore not to hire data scientists. **Argument 2 — skills mismatch**: what is mainly needed is **development and integration engineers** (back/front), **strong cloud skills**, and **DevOps**. Client quote: *« You don't necessarily need to be a data scientist, but you need to understand the basic concepts, have back-office development skills, and strong cloud skills. »* **Argument 3 — platform architecture (orchestrators + APIs)**: building an enterprise **plateforme d'IA générative** via orchestrators and APIs makes it *« possible to work with the best LLMs on the market and switch between them as their respective capabilities evolve, without reworking the applications »* (anti vendor lock-in). **Argument 4 — from project to product**: *« The platform […] must be regarded as a product in its own right »*; instead of a one-off investment, plan for a **monthly funding stream** (continuous iteration, ongoing innovation). **Argument 5 — governance & shadow AI**: the unprecedented democratization of GenAI generates *« as much shadow AI as strong expectations toward the CIO office »* → governance to capture business needs, **prioritize products by value**, and oversee proper operation. **Paradigm shift** announced: *« the shift is from classic algorithmic programming to agents Langchain that handle part of the decisions »*. **Relevance to the watch**: a **founding text (2 years ahead)** of WeNvision's doctrine (product > project, platform/API, flow-based funding, governance, shadow AI), later extended by [[wenvision-ai-agents-enterprise-deployment-2025-10-01]], [[habert-ia-agentique-production-2025-10-29]], and rafal-wenvision-tokenomics-foundation-finops-ia-2026-06-04 (FinOps/token, flow-based funding → financial governance). It also foreshadows the *harness/platform around the model* (Dropbox/Okumura: *systems around the model*) and **model independence** achieved through an orchestration layer.

## Titre Article

L'IA générative est plus une affaire de produit technologique qu'un projet d'IA

## Date

2024-02-23

## URL

https://www.cio-online.com/actualites/lire-l-ia-generative-est-plus-une-affaire-de-produit-technologique-qu-un-projet-d-ia-15485.html

## Keywords

generative AI, technology product, product vs project, data science, foundation model, base model, millions of euros, Bloomberg, BloombergGPT, data scientist, integration engineers, back-office development, cloud skills, DevOps, plateforme IA générative, orchestrators, API, model independence, vendor lock-in, switching LLMs without reworking applications, product management, flow-based funding, monthly stream, continuous iteration, governance, value-based prioritization, shadow AI, expectations toward the CIO office, democratization, algorithmic programming, agents Langchain, paradigm shift, Olivier Rafal, WeNvision, CIO-Online, CIO office

## Authors

**Olivier Rafal**, *Consulting Director Strategy* chez **WeNvision** (cabinet de conseil FR). Tribune publiée dans la rubrique *Tribune* de **CIO-Online**. Auteur déjà présent dans la veille (cf. fiches WeNvision/Atlas/Tokenomics). Publié le **23 février 2024**.

## Ton

**Profile**: Expert-advisor op-ed (CIO *thought leadership*), addressed to **CIOs and IT decision-makers**, **educational-prescriptive** register, moderate technical level (architecture and product-management notions accessible to executives). Insider perspective familiar with client projects (cites a client testimonial).

**Style**: Funnel-shaped demonstration — first deconstructs a received idea ("you need data scientists"), then reconstructs the right approach (integration skills + platform + product + governance). Draws on **concrete examples** (Bloomberg/BloombergGPT), a **client quote**, and operational notions (orchestrators, APIs, funding stream). Measured tone, no hype, oriented toward **real-world implementation** on the CIO side. An early text (Feb. 2024) that anticipates theses that have since become mainstream.

**Key aphorisms**:
- ***« L'IA générative est plus une affaire de produit technologique qu'un projet d'IA. »*** (title-thesis).
- *« Building a foundation model from scratch requires several months, millions of euros, and access to enormous quantities of data. »*
- *« The platform […] must be regarded as a product in its own right. »*
- *« Work with the best LLMs on the market and […] switch between them […] without reworking the applications. »*
- *« The shift is from classic algorithmic programming to agents Langchain that handle part of the decisions. »*

**Metaphors / frameworks at work**:
- ***Product, not project*** — rejection of the "project with a defined end + one-off investment" model in favor of a product funded in **flow** and iterated continuously.
- ***Platform as abstraction layer*** — orchestrators + APIs = independence from models (anti lock-in).
- ***Skills mismatch*** — from the data scientist toward the integration / cloud / DevOps engineer.
- ***Shadow AI vs. governance*** — democratization creates expectations and risks that the CIO office must channel through governance and value-based prioritization.

**Epistemic position**: an argued opinion op-ed (advisory), grounded in field experience and an emblematic example (Bloomberg). No proprietary quantitative data; value = a **CIO decision framework** laid out early. Best read as a **dated doctrinal milestone** (Feb. 2024) rather than a study.

**Authority**: (a) position as **strategy consulting director** (WeNvision); (b) **precedence** of the argument (product > project thesis laid out as early as 2024); (c) **concrete grounding** (Bloomberg, client quote); but (d) op-ed format, without proprietary metrics.

## Pense-betes

- **Date / source**: **February 23, 2024**, **CIO-Online** (Tribune). Author: **Olivier Rafal**, Consulting Director Strategy **WeNvision**. **Founding text** (≈2 years before the agentic WeNvision fiches).
- **Thesis**: ***« L'IA générative est plus une affaire de produit technologique qu'un projet d'IA »*** → don't turn it into a data science project. ### The 5 messages 1. **No need for data scientists (in general)**: a *foundation model* costs *« several months, millions of euros »* + massive data → reserved for players with specific datasets (**Bloomberg → BloombergGPT**). 2. **Key skills**: **development/integration** engineers (back/front), **strong cloud**, **DevOps**. Client quote: *« don't necessarily need to be a data scientist […] strong cloud skills. »* 3. **Platform = orchestrators + APIs**: *« switch [LLMs] […] without reworking the applications »* (model independence, anti lock-in). 4. **Product > project**: *« regard [the platform] itself as a product »* + **monthly flow-based funding** (vs one-off investment). 5. **Governance & shadow AI**: unprecedented democratization → *« shadow AI »* + *« strong expectations toward the CIO office »*; capture needs, **prioritize by value**. ### Sign of the times
- **Paradigm shift** already announced: *« from classic algorithmic programming to agents Langchain that handle part of the decisions »* (Feb. 2024). ### To use in engagements / presentations
- **WeNvision doctrine milestone**: lays out as early as 2024 *product > project*, *platform/API*, *flow-based funding*, *governance*, *shadow AI* — a throughline up to the **financial governance / FinOps token** of the Tokenomics Foundation op-ed (June 2026).
- Argument reusable on the CIO side: **model independence** via an orchestration layer (foreshadows *systems around the model*).
- **Flow-based funding** (2024) → conceptually anticipates the **FinOps / recurring per-token cost** logic (2026).

## RésuméDe400mots

In this op-ed published on **February 23, 2024** on **CIO-Online**, **Olivier Rafal** (Consulting Director Strategy at **WeNvision**) defends an idea that was then against the grain: **generative AI is more a matter of technology product than an AI or data science project**. Many organizations misjudge their priorities by seeking to hire data scientists and machine learning engineers.

First argument: building a *foundation model* from scratch *« requires several months, millions of euros, and access to enormous quantities of data »*. This only makes sense for players with specific, monetizable datasets — the emblematic example being **Bloomberg**, which created **BloombergGPT** to leverage its financial data. For nearly all companies, it is better to rely on commercial LLMs.

Second argument: the skills mismatch. Rather than data scientists, what is needed is **development and integration engineers** (back and front), **strong cloud skills**, and **DevOps**. A client sums it up: *« You don't necessarily need to be a data scientist, but you need to understand the basic concepts, have back-office development skills, and strong cloud skills. »*

Third argument: architecture. Companies should build a **plateforme d'IA générative** based on **orchestrators and APIs**, which makes it *« easy to work with the best LLMs on the market and switch between them as their respective capabilities evolve, without reworking the applications »* — a guarantee of independence against *vendor lock-in*.

Fourth argument, the central one: the shift **from project to product**. *« The platform […] must be regarded as a product in its own right »*, funded not by a one-off investment but by a **monthly stream**, to sustain continuous iteration and innovation.

Fifth argument: **governance**. GenAI *« has become democratized in an unprecedented way »*, which generates *« as much shadow AI as strong expectations toward the CIO office »*. Adequate governance must capture the needs of the various business units, **prioritize products by the value** they create, and oversee the proper functioning of the whole.

Rafal finally announces a **paradigm shift** that must be accepted: *« the shift is from classic algorithmic programming to agents Langchain that handle part of the decisions »*. A founding text, it lays out as early as 2024 the foundation (product, platform/API, flow-based funding, governance) that later analyses — up to the 2026 FinOps/token financial governance — would only extend.

## GrapheDeConnaissance

- Olivier Rafal —publie→ L'IA générative est plus une affaire de produit technologique qu'un projet d'IA (DOCUMENT, 0.97)
- Olivier Rafal —travaille_chez→ WeNvision (ORGANISATION, 0.95)
- CIO-Online —publie→ L'IA générative est plus une affaire de produit technologique qu'un projet d'IA (DOCUMENT, 0.95)
- Olivier Rafal —affirme_que→ l'IA générative est un produit technologique plus qu'un projet d'IA (AFFIRMATION, 0.96)
- Olivier Rafal —affirme_que→ créer un foundation model demande plusieurs mois et des millions d'euros (AFFIRMATION, 0.93)
- Bloomberg —a_créé→ BloombergGPT (TECHNOLOGIE, 0.95)
- Olivier Rafal —recommande→ ingénieurs d'intégration et compétences cloud plutôt que data scientists (CONCEPT, 0.9)
- plateforme d'IA générative —est_basé_sur→ orchestrateurs et API (TECHNOLOGIE, 0.92)
- orchestrateurs et API —permet→ de changer de LLM sans retoucher les applications (CONCEPT, 0.92)
- Olivier Rafal —recommande→ considérer la plateforme GenAI elle-même comme un produit (AFFIRMATION, 0.94)
- Olivier Rafal —recommande→ financement en flux (CONCEPT, 0.9)
- démocratisation de la GenAI —permet→ shadow AI (CONCEPT, 0.9)
- gouvernance GenAI —utilise→ priorisation des produits par la valeur créée (CONCEPT, 0.88)
- agents Langchain —remplace→ programmation algorithmique classique (METHODOLOGIE, 0.88)

---
Canonical: https://www.thekb.eu/en/fiches/rafal-wenvision-ia-generative-produit-techno-pas-projet-2024-02-23/