AeronauticsINTELLIGENCE ARTIFICIELLE
100% coverage, half the effort: AI transforms supplier onboarding review
ALTEN in Morocco built a fine-tuned LLM assistant for a leading global automotive manufacturer, automating the review of Supplier Onboarding Managers (SOM) entries to reduce verification effort, improve data consistency, and achieve full quality coverage across supplier readiness processes.
As automotive manufacturers collaborate with large supplier ecosystems, onboarding and logistics readiness activities generate significant volumes of information that must be reviewed and validated before production milestones. Engineers and project leaders spend considerable time assessing supplier information, checking data consistency, and ensuring compliance with onboarding requirements. These manual verification activities are resource-intensive and prone to inconsistencies.
A leading global automotive manufacturer partnered with ALTEN to develop an AI-powered assistant capable of reviewing and validating supplier onboarding records at scale, automating the assessment of onboarding status, planned actions, blockers, and project milestones while improving data quality and process efficiency.
Challenge: Ensure the quality and consistency of Supplier Onboarding Management (SOM) data by automatically reviewing supplier entries and records, reducing verification effort, and achieving complete quality coverage across readiness activities.
Solutions: A fine-tuned LLM assistant trained on onboarding instructions, templates, and SOM data to assess comments, validate consistency, identify anomalies, and recommend corrections, processing approximately 3,300 SOM entries per week.
Benefits:
- 50% reduction in verification effort for SOM team and Project Leaders
- 100% coverage of SOM Entries reviews & verification
- Improved quality with Right First Time (RFT) reaching 99%
- Fewer iterations between SOM team and project leaders
A fine-tuned assistant for supplier onboarding management
ALTEN built a fine-tuned LLM assistant using onboarding instructions, reference templates, and SOM (Supplier Onboarding Management) files to review supplier entries at scale. The assistant analyses entries, evaluates their consistency, detects anomalies, and proposes corrections automatically. The assistant significantly reduces the effort required to maintain high-quality onboarding data.
Delivered using PyTorch, Python, NLP technologies, and automation components, integrating seamlessly into existing supplier readiness workflows.
Faster verification, better onboarding quality
The assistant allows supplier onboarding teams to focus on higher-value support activities rather than repetitive verification tasks. Beyond productivity gains, the solution provides a scalable way to maintain consistent, high-quality onboarding process while reducing operational effort.
The power of AI has enabled this automotive manufacturer to review supplier onboarding data faster, more accurately, and at a scale manual verification could not sustain, ensuring complete quality coverage while significantly reducing verification effort.
