Vendor Tier Classification Workflow architecture diagram

About This Architecture

Vendor Tier Classification Workflow orchestrates real-time vendor assessment by routing adjuster requests through a central classifier that integrates structured vendor profiles, historical evaluation data, and local LLM inference. The Vendor Tier Classifier queries vendor_profiles.json and vendor_eval_golden.json for baseline data, retrieves contextual vendor history and trade patterns from ChromaDB Vector Store, and invokes Ollama LLM Server (gemma2:2b and gemma4:e4b models) for intelligent tier recommendations. This architecture enables procurement teams to classify vendors consistently without external API dependencies, maintaining data privacy while leveraging vector similarity search and generative AI. Fork and customize this workflow on Diagrams.so to adapt tier criteria, swap LLM models, or integrate additional data sources.

People also ask

How do I build a vendor classification system using local LLMs and vector search without cloud APIs?

This diagram shows a Vendor Tier Classifier that retrieves vendor profiles and evaluation history from JSON files and ChromaDB, then queries Ollama LLM (gemma2:2b or gemma4:e4b) to generate tier recommendations. The workflow maintains data privacy by running inference locally while leveraging vector similarity for contextual vendor history and trade patterns.

Vendor Tier Classification Workflow

AutointermediateLLMChromaDBOllamavendor-managementprocurementvector-search
Domain: Ml PipelineAudience: ML engineers and procurement specialists building vendor evaluation systems with local LLMs
5 views0 favoritesPublic

Created by

July 13, 2026

Updated

August 18, 2026 at 10:33 AM

Type

flowchart

Need a custom architecture diagram?

Describe your architecture in plain English and get a production-ready Draw.io diagram in seconds. Works for AWS, Azure, GCP, Kubernetes, and more.

Generate with AI

AI-generated. Verify before production use. Learn more

Report this diagram