AI agent deployment for customer service

How OpenEd Auto-Resolved 70% of Support Chats in Four Weeks

See how fwdDeploy deployed a Fin AI - (Formerly Intercom) for OpenEd, auto-resolving 70% of chats in four weeks. FDE-led, live in weeks.

CUSTOMER

OpenEd
(formerly My Tech High)

INDUSTRY

Education
(online & blended learning)

SCALE

25,000+ students across school partnerships

SOLUTION

An AI support agent on Fin AI - (Formerly Intercom), deployed by fwdDeploy

OpenEd results: Fin AI handled 474 chats in 4 weeks; 70% deflected, 67.3% resolved, 96.4% answered

Results at a Glance

70%

AUTO-RESOLVED

96.4%

AI ENGAGEMENT

4 weeks

TO FULL IMPACT

At a Glance

Customer

OpenEd (formerly My Tech High)

Industry

Education (online and blended learning)

Scale

25,000+ students across school partnerships

Challenge

A manual support stack that could only grow by hiring

Solution

An AI support agent on Fin AI - (Formerly Intercom), deployed by fwdDeploy

Outcome

70% of chats auto-resolved. Full AI impact in four weeks

Challenge

Support that is scaled only by hiring

OpenEd serves more than 25,000 students through school partnerships. Every enrollment period turns that reach into a wave of questions. A 30-person team handled over 50,000 queries a year across phone, email, and chat. There was no way to keep up except to hire.

The legacy Zendesk stack made it harder. It had little automation and no self-service. Agents answered the same routine questions by hand every day. Conversations sat across disconnected channels. Families waited. Every spike in demand became a spike in workload.

OpenEd did not need a bigger team. They needed an AI agent doing the front-line work.

Who can deploy an AI agent

OpenEd wanted a partner who can dpeloy AI agent for their customer support. The goal was to unify their channels, automate the routine volume, and stand up an AI agent that reflected how their support actually ran. fwdDeploy built it on Fin AI (formerly Intercom). 

How We Did It

Discover, Engineer, Optimize

fwdDeploy runs every deployment in three phases. OpenEd went live in four weeks.

Discover

We mapped OpenEd's workflows, user journeys, and pain points. We ranked the highest-volume query types in a prioritization matrix. That told us what the AI agent should resolve first.

Engineer

We configured Intercom end to end: channel routing, workflows, bots, rules, and escalation paths. We migrated workflows and data off Zendesk. We grew the knowledge base from fewer than 100 articles to more than 300. We integrated Intercom with OpenEd's internal systems. Then we configured Fin AI to resolve common questions and hand complex cases to agents with full context.

Optimize

We trained the OpenEd team on bot tuning, campaign settings, reporting, and governance. We stayed on to tune the agent in production against real resolution metrics.

Results

70% deflection in the first four weeks

OpenEd's 50,000+ annual queries once demanded constant manual effort. After launch, Fin AI - (formerly Intercom) took the front line of chat support. In the first four weeks, it handled 474 chat conversations. It resolved most of them with no human involved.

OpenEd Fin AI results: 96.4% answered, 70% deflected (332/474), 67.3% resolved, 30% escalated, 2.7% abandoned

Agents once spent their days on routine questions. Now they focus on the 30% of conversations that need human judgment. OpenEd got there in weeks. 

Scale your support the same way

Ready to resolve most of your support volume automatically?

fwdDeploy deploys Fin AI (formerly Intercom) agent for customer service. We cover discovery, build, migration, and ongoing optimization. Go-live and get value in weeks.

Book a diagnostic call and put your support on the same footing as OpenEd.

Frequently Asked Questions

How did OpenEd auto-resolve 70% of its support chats?

How long does it take to deploy an AI support agent?

What is the difference between deflection rate and resolution rate?

Does an AI support agent replace human agents?

Can you migrate from Zendesk to Fin Intercom without disrupting support?

How does fwdDeploy deploy AI support agents?

Your results next

Get results like these for your service stack.

One call. No lengthy discovery cycle. Our forward-deployed engineer maps what AI can take off your team's plate.