Case Studies | SaaS B2B

Support that verifies first

Customer Support Agent

About

A SaaS B2B business working with Cloud Combinator on AWS. The client is anonymised at their request.

Challenge

The challenge had four focus areas.

Handling first-line support at scale

Customers ask the same kinds of questions again and again: where is my booking, what do I owe, when is my drop-off. Answering those instantly on the website, without tying up a human agent, was the core opportunity.

Verify before you share

A rental account holds personal and payment data, so the agent can never volunteer information to an unverified caller. The bot has to look a booking up, then ask a verification question, and only share details once the answer checks out, with GDPR compliance meaning no personal data is exposed before verification passes.

Multi-tenancy with strict isolation

The same agent has to serve many rental operators while returning only the data that belongs to the right tenant. At the client's scale, a leak of one operator's data into another operator's conversation would be far more damaging than a slow answer.

Knowing when to escalate

The agent is deliberately read-only for the proof of concept and cannot resolve everything. When it cannot help, or when verification fails, it needs to hand off gracefully to a human rather than guess or stall.

Solution

The engagement ran as a four-week proof of concept, moving from a working skeleton to a tested, demonstrated agent against the client's own demo environment.

100%

Verify-before-share; no personal data released without successful verification

1 bot

Multi-tenant from day one, serving every rental operator with isolated data

4 wks

From kick-off to demo and handover

By the numbers:

  • 100% - Verify-before-share; no personal data released without successful verification
  • 1 bot - Multi-tenant from day one, serving every rental operator with isolated data
  • 4 wks - From kick-off to demo and handover
Changes

Acceptance is defined against safe, correct behaviour: the bot classifies intents correctly, never shares data before verification passes, fails gracefully and escalates, and returns data only for the correct tenant. The figures below are the targets and design guarantees set out in the Statement of Work.

  • Instant first-line answersThe agent resolves the common questions, booking status, invoice details, payment history and vehicle information, without a human in the loop.
  • Safety firstIdentity verification is a hard gate: the bot shares nothing until the caller passes, keeping the experience GDPR-compliant by design.
  • True multi-tenancyA single bot serves all operators, reusing the client's existing header-based isolation so each customer only ever sees their own data.
  • Graceful escalationWhen the agent cannot help, or verification fails after two attempts, it hands off cleanly to the client's support team by email rather than guessing.
  • HandoverThe proof of concept closes with a final demo, documentation, source code and a clear production recommendation covering website integration and the optional Bedrock enhancement.

With a safe, multi-tenant agent proven against their own environment, the client finishes the proof of concept with a clear path to production, ready to take first-line support off their operators' hands and to layer in richer, more natural responses whenever they choose.

AWS Stack

Amazon Lex V2

For intent classification, slot collection and dialogue management.

AWS Lambda

For fulfilment logic, identity verification and calls to the client's REST API.

Amazon Bedrock

(Claude 3 Haiku) for optional natural, tone-aware response generation.

Amazon SES

For escalation emails to the client's support team when the agent cannot resolve a query.

Amazon Cognito

And AWS CloudFormation for secure deployment and browser access to the chat widget.

YOU MIGHT LIKE

Related success stories

View all case studies

Case Studies | Insights

Utilising Language Recognition, Speed, and Enhanced Security to Make Social Media a Force for Good

  • Here, we take a detailed look at how the Cloud Combinator team collaborated with another cutting-edge AI service provider that provides intelligent systems to “make social media more social” for brands and users alike.
  • Arwen AI is a UK-based startup specialising in AI solutions to manage and enhance brands’ social media interactions. Founded in 2020 by Matt McGrory, Dr. David Cole, and Joel Bailey, Arwen. AI focuses on using AI to automatically detect and remove spam, toxic comments, and other unwanted content from social media platforms.
  • The team at Arwen have three core products. ‘Moderate’ is focused on identifying and removing toxic content from social media channels. ‘Engage’ helps brands identify and engage with meaningful conversations on social media, and ‘Customize’ allows brands to apply bespoke algorithms to their channels - creating an even more effective moderation and engagement.
Read more
CONTACT US

Ready to turn AI into impact?

We'll help you spot the highest-value opportunities, reduce risk around your first AI initiative, and define a clear path to results from day one.

Why talk to us:

Outcome-driven recommendations

AWS-recognised delivery expertise

Risk-aware AI adoption

Clear next step, not a sales pitch

Start with a focused 20-minute conversation about your goals — no pressure, no commitment.

This website uses cookies to enhance user experience and to analyze performance and traffic on our website.

See our Privacy Policy for details.