Agent 18 · E-Commerce
EmailFlow Agent
Recovers abandoned carts with consent-checked emails.
Agent PackCommerceLangGraph + CrewAI
Found25
Fixed + verified25
Held for approval0
Recall1.00
One offline run on demo data with planted problems (recorded platform test, November 2026).
How EmailFlow heals data, step by step
INData inSynthetic carts
01Detect25 problems found
02Diagnose2 problem types named
03Plan25 safe to fix · 0 need a person
04Fix + verify25 fixed and re-checked
05ReportPlain-English report and Control Tower record
- Client problem
- Carts are abandoned and invalid emails waste sends.
- What it does
- Validates emails and consent, finds abandoned carts.
- How it heals
- Suppresses invalid addresses and sends personalised recovery emails (CrewAI copy) only to customers with consent who have not unsubscribed.
- Human gate
- No email without consent.
- Typical buyer
- D2C brands
- Built with
- LangGraph self-healing loop with a CrewAI crew (analyst, fixer, reviewer)
- Demo data
- Synthetic carts (shaped on the online-retail customers).
- Tables
- carts ← synthetic
- Target outcome
- Target: 5-15% cart recovery.
- Path to production
- Commerce Pack.
Issues detected on the demo data
| Issue | Risk | Proposed fix | Count |
|---|---|---|---|
| bad_format_email email='gargyash at gmail.com' fails format check | low | Suppress invalid address from all sends | 5 |
| cart_abandoned Cart worth 7042 abandoned 37h ago | low | Send personalised recovery email | 20 |
How we demo it
- Load carts.
- Run EmailFlow.
- Open the emails in Mailpit.
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.