Inetum

Industries and client fit

Documented engagements include public-sector procurement and education administration, a hospital IT helpdesk, and railway platform services. These provide concrete reference contexts for buyers assessing administrative workflow automation and enterprise IT delivery; they do not establish experience with every clinical or sector-specific process. ^{[12] [13] [4] [10]}

Engagement and delivery

Inetum describes local, nearshore, and offshore delivery capabilities. Its application practice explicitly offers build-only, run-only, and build-and-run contracts. Buyers should confirm the proposed delivery locations, division of responsibilities, governance, and operational commitments for their contract. ^{[1] [2]}

Track record

For Est Ensemble, Inetum describes developing and operating an Azure OpenAI procurement tool that checks documents, evaluates bids against requirements, extracts supporting passages, and proposes scores. Inetum reports approximately 30% less processing time per file; this is a case-specific reported result, not a general performance commitment. ^[12]

At UZ Gent, the destination case study describes an IT-helpdesk chatbot built using Microsoft Azure AI Foundry, grounded in validated internal knowledge and returning source links. Inetum describes the framework as production-ready and embedded in daily operations. This evidence concerns IT support, not clinical summarization. ^[4]

For Spain’s education ministry, Inetum describes document extraction and business-rule validation using Microsoft Azure and Azure OpenAI, including proof-of-concept testing and prompt refinement. A separate INTEF account describes GenAI EDU combining contextual analysis with Azure Document Intelligence while retaining human decision authority. ^{[14] [13]}

Scope and limitations

• The supplied evidence does not establish public pricing, standard project sizes, contractual SLAs, or the delivery team assigned to a particular customer.

• AI delivery evidence is strongest for generative AI. Comparable detail for custom conventional machine-learning model development was not established.

• Coborg and GenAI EDU licensing, customization boundaries, third-party charges, and maintenance responsibilities are not clearly separated in the supplied evidence.

• The Spanish education-ministry account does not disclose production acceptance criteria or long-term support arrangements. Planned AI work in the SNCF account is not evidence of completed AI deployment.

Questions to ask

• Which deliverables are bespoke, which use Inetum proprietary assets, and which require third-party licenses?

• Who owns prompts, configurations, integrations, evaluation datasets, and generated artifacts?

• What accuracy, grounding, security, and human-review tests govern AI acceptance?

• Who operates the system after launch, including model changes, incident response, and recovery?

• Which delivery locations and teams are committed, and how are implementation and operational responsibilities divided?

• Which references match our platform, regulatory environment, and required delivery stage?