Agent 17 · E-Commerce
RecoEngine Agent
Fills recommendation gaps for new products.
Narrow scope firstCommerceLangGraph
Found5
Fixed + verified5
Held for approval0
Recalln/a
One offline run on demo data with planted problems (recorded platform test, November 2026).
How RecoEngine heals data, step by step
INData inKaggle carrie1/ecommerce-data invoice history.
01Detect5 problems found
02Diagnose1 problem type named
03Plan5 safe to fix · 0 need a person
04Fix + verify5 fixed and re-checked
05ReportPlain-English report and Control Tower record
- Client problem
- New products never get recommended (cold start).
- What it does
- Builds co-purchase recommendations and finds products with none.
- How it heals
- Fills cold-start items with content-based picks.
- Human gate
- None.
- Typical buyer
- Larger online stores
- Built with
- LangGraph self-healing loop
- Demo data
- Kaggle carrie1/ecommerce-data invoice history.
- Tables
- products ← Kaggle PromptCloudHQ/flipkart-products
- Target outcome
- Target: 100% of active products with recommendations.
- Path to production
- Later, once the client has traffic.
Issues detected on the demo data
| Issue | Risk | Proposed fix | Count |
|---|---|---|---|
| cold_start 84 catalogue items have no recommendations | low | Generate content-based recommendations | 5 |
How we demo it
- Run RecoEngine.
- Show a new product now with recommendations.
Built on AWS
Runs in the platform container on EC2 Graviton (or on demand in Lambda), writes healed tables to the S3 lakehouse, uses Amazon Bedrock only to write explanations, and reports six metrics to CloudWatch.
Step into DigiQuench.
Agents that heal your data.
Watch an agent find planted problems, fix the safe ones and wait for your approval on the rest. Thirty minutes, live.