ecommerce_ops · ecommerce · workflow
Mercado Libre builds multi-LLM orchestration pipeline for product matching at 95% precision and sub-$0.001 cost per request
Mercado Libre's product-matching pipeline required manual human oversight for high-stakes use cases such as pricing because the existing ML model was not accurate enough to operate fully autonomously.
How it works
Common implementation structure
How this type of workflow is generally built, generalized across documented cases — not tied to any one vendor's stack. Click any stage to read what happens there. Specific products that implement these stages appear in “Tools commonly seen” below.
Stage 1 · Select target item
A specific item requiring matching is selected as the starting point of the pipeline.
Tools used
VerdiANNvector databasesembedding models
Outcome
After iterative orchestration design and prompt engineering, Mercado Libre achieved 95% precision and at least 50% recall at a cost of less than $0.001 per request, reaching human-level performance with an autonomous method scalable across millions of items.
What failed first
An initial three-step LLM pipeline produced only 38% precision and 13% recall, far below the existing ML model's performance. Even the sophisticated ML model combining advanced NLP, multiple embeddings, and attribute enrichment could not fully automate the process.
Results
Volume38%
Cost replacedless than $0.001
Grounding & classification
Source type: technical build writeup
30 fields verified against source quotes.
agentic workflowmulti agent workflowpredictive analyticsproduct catalogfailure mode describedhuman review describedmetric backednamed customerproduction runtime claimedtools describedworkflow describedecommerceaccuracy improvementautomation ratecost reductiontechnical build writeupecommerce opsextract classify routehuman review queue