Customer support · Production

Consensys scales secure Web3 support with Intercom Fin AI Agent, resolving over 70% of conversations

The problem

Consensys served millions of pseudonymous global customers managing tokens, wallets, and financial access at high stakes, while slow support created exploitable windows for scammers. Their legacy two-vendor stack for ticketing and live chat was too slow, expensive, and rigid to iterate fast enough to keep users safe.

First attempt

Consensys's legacy stack relied on two established vendors — one for ticketing, one for live chat — with heavy dependence on vendor-led bots and professional services that made every configuration change an external project, leaving the team unable to own or rapidly adapt their automation stack.

Workflow diagram · grounded in source
1
Customer contacts support
trigger
“Fin deployed for core channels like Messenger and email”
2
Fin AI resolves conversation
ai_action
“Fin now handles nearly 20,000 monthly resolutions with 90% involvement in conversations and and a resolution rate of over 70% that continues to grow”
3
Real-time translation
ai_action
“Real-time translation capabilities now serve users across nearly 200 countries in dozens of languages”
4
Complex queries routed to humans
routing
“human agents are freed up for high-touch, high-context queries”
5
Fast resolution closes scammer window
output
“Faster, more accurate responses with 24/7 availability help close the critical "window of opportunity" that scammers exploit when legitimate support is unavailable or delayed”
Reported outcome

Within 8 weeks Fin was resolving nearly 70% of support conversations with matching CSAT.
Today Fin handles approximately 20,000 resolutions per month with 90% involvement in conversations and a resolution rate over 70%, while real-time translation serves users across nearly 200 countries, enabling a true AI-first support organization.

Reported metrics
Resolution rate within 8 weeksnearly 70%
Monthly resolutions~20,000
Resolution rate vs competitors20%
CSAT vs competitors15%
Show all 10 reported metrics
resolution rate within 8 weeksnearly 70%
monthly resolutions~20,000
resolution rate vs competitors20%
CSAT vs competitors15%
involvement in conversations90%
current resolution rateover 70%
countries servednearly 200
early containment rate50%
CSATwith the CSAT to match
customer reviewsCustomers routinely praise Fin in 5-star reviews
Reported stack
FinIntercomMessenger
Source
https://www.intercom.com/customers/consensys
Read source ↗

Frequently asked questions

What did this team achieve with this AI workflow?

Within 8 weeks Fin was resolving nearly 70% of support conversations with matching CSAT.

What tools did this team use?

Fin, Intercom, Messenger.

What results were reported?

Resolution rate within 8 weeks: nearly 70%; Monthly resolutions: ~20,000; Resolution rate vs competitors: 20%; CSAT vs competitors: 15% (source-reported, not independently verified).

What failed first in this deployment?

Consensys's legacy stack relied on two established vendors — one for ticketing, one for live chat — with heavy dependence on vendor-led bots and professional services that made every configuration change an external p…

How is this customer support AI workflow structured?

Customer contacts support → Fin AI resolves conversation → Real-time translation → Complex queries routed to humans → Fast resolution closes scammer window.