AI ChatbotsE-commerce

Glimr

78% deflection, CSAT up to 4.6/5

Support tickets were burying a 12-person team.

Claude 4PineconeShopify APISlack APINext.js

Tier-1 deflection

78%0%

+78%

Avg response time

0.8s6.2 hrs

+99%

CSAT score

4.6/53.2/5

+44%

Support team size

312

+75%

Challenge

Glimr's support team was drowning in tier-1 tickets: order lookups, return policies, shipping questions. Response time was 6+ hours. CSAT was 3.2/5. They had three SaaS subscriptions just for support tools.

Solution

We deployed a RAG-powered AI chatbot trained on their docs, policies, and order API. It handles 78% of tier-1 queries autonomously, escalates with full context, and integrates with their Shopify + Slack stack.

Outcome

78% deflection, CSAT up to 4.6/5. The results speak for themselves — every metric moved in the right direction, and the client saw ROI within the first month of launch.

Project timeline

Week 1

Knowledge audit

Indexed docs, FAQs, policies, order data schema

Week 2

Build + eval

Bot build, 50-question eval against ground truth

Week 3–4

Pilot deploy

Soft launch, transcript review, iteration

Week 5+

Scale

Add channels, automate evals, monthly reviews

“It deflects 78% of tier-1 tickets and our CSAT actually went up. We canceled three SaaS subscriptions.”

Adam Nakai

Head of Support, Glimr

Listen to what they said

Live dashboard

Analytics

Live

12.4K

+24%

$84K

+18%

8.7%

+312%

3.2K

+45%

Weekly traffic

RAGClaudeShopifySlack