Case Studies | SaaS B2B

Getting agentic AI past the security gate

BRONTOWATCH AI Security Posture Accelerator

About

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

Challenge

The feasibility work confirmed a buyer pain that is real and consistent across regulated startups and ISVs. It breaks down into three focus areas.

Stuck at the security gate

Agentic AI projects stall between pilot and production because the CISO has no visibility over what agents reason, which tools they call, what identity they assume, and what data they touch. With nothing to see, approval defaults to no.

A new threat class existing tooling misses

Prompt injection, tool abuse, runaway loops, data leakage through agent reasoning, and identity confusion are specific to the agent decision layer. Traditional security stacks watch infrastructure and network, not how an agent decides and acts.

No audit evidence, and retention too short to help

Security and compliance teams cannot produce the traceability and attestation their reviews demand, and native log retention windows are too short for meaningful forensics across an agentic estate. Without durable, correlated evidence, sign-off is impossible.

Solution

The partners packaged their combined capability into a fixed-scope, twelve-week accelerator. BRONTOWATCH observes, correlates and reports only; it takes no write actions against production systems, which keeps the operational risk of the delivery minimal. Cloud Combinator leads delivery on AWS with the client as the telemetry data plane. Delivery runs in four stages.

12 weeks

Fixed-scope accelerator from stalled pilot to production sign-off

12 months

Always-hot, correlated agent audit trail retained for assurance

~1 day

To stand up the telemetry pipeline via the reusable deployment kit

By the numbers:

  • 12 weeks - Fixed-scope accelerator from stalled pilot to production sign-off
  • 12 months - Always-hot, correlated agent audit trail retained for assurance
  • ~1 day - To stand up the telemetry pipeline via the reusable deployment kit
Changes

The engagement produced a repeatable, ship-to-production security programme rather than an one-off assessment, validated for delivery through the BOX programme with a reusable deployment kit that stands the pipeline up in about a day.

  • Visibility over the agent layer — Reasoning, tool calls, identity, and data flows are captured and correlated end to end, so security teams can see what agents actually do rather than assume.
  • Audit-grade by default — A durable, queryable evidence record turns the CISO conversation from assertions into proof, unblocking production sign-off.
  • Delivered outcome, not raw tooling — AWS ships the telemetry primitives and the client ships the data plane; BRONTOWATCH turns them into a correlated evidence record with an agentic detection-rule library.
  • A minimal risk surface — The read, correlate and report design takes no write actions against production, so the delivery itself adds no operational exposure.
  • Handover and recurring assurance — Enablement, runbooks and a support model at handover, with BRONTOWATCH Watch offering quarterly detection-rule updates and posture reviews as the estate grows (projected expansion offer).

With the solution defined and the deployment kit proven repeatable, the partnership moves to securing its first design-partner pilots in FinTech and healthcare, turning a reference architecture for securing agentic AI into live customers and a scalable co-sell motion with AWS.

AWS Stack

Amazon Bedrock

And Amazon Bedrock AgentCore (Runtime, Identity, Observability, Gateway, Memory) for the agent runtime and native telemetry.

AWS Distro

For OpenTelemetry (ADOT) for standards-based instrumentation of the agent layer.

AWS CloudTrail

And Amazon CloudWatch for identity, data-access and invocation logging.

Amazon Data Firehose

And AWS Lambda for forwarding logs into the client's pipeline.

Amazon S3

For evidence and log storage underpinning the audit trail.

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.

We use cookies

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

See our Privacy Policy for details.