Case Studies | SaaS B2B

Auto-populating digital deal rooms with generative AI

LLM Deal Room Widget Population

About

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

Challenge

The engagement had three focus areas.

Reading messy sales decks reliably

Sales decks vary in layout and length. The pipeline needed to accept many decks at once, uploaded securely, and extract the information inside them accurately and in a timely way, ready for a language model to reason over.

Answering the right questions accurately

The value is in specific answers: three things about the company, features offered, why a customer would choose them, the people involved and their roles. The solution had to feed the extracted content to a Bedrock model and return dependable, structured answers for each widget.

Doing it securely and within cost

Because untrusted documents drive the prompt, the pipeline had to guard against attempts to hijack the model, keep prompts within the model's context window, alert when token usage ran high, and be deployable as secure, repeatable infrastructure inside the client's own AWS environment.

Solution

Delivered under Cloud Combinator's AI and ML Project Accelerator, the work moved from a secure foundation through extraction to answer generation, tested against the client's own example decks.

Textract +

Document extraction feeding a large language model

4,000/mo

Deal room uses the pipeline is costed to support (projected)

JSON

Structured question and answer output for each widget

By the numbers:

  • Textract + - Document extraction feeding a large language model
  • 4,000/mo - Deal room uses the pipeline is costed to support (projected)
  • JSON - Structured question and answer output for each widget
Changes

The proof of concept demonstrated the full path, from many uploaded decks through timely Textract extraction to reliable, structured answers from Bedrock, proving that AWS generative AI can populate the client's widgets automatically.

  • Extraction proven — Many sales decks uploaded securely via presigned URL and processed by Amazon Textract in a timely manner.
  • Accurate, structured answers — An Amazon Bedrock model answering the client's defined questions from the extracted content, returned as clean key value pairs.
  • Secure by design — Guards against prompt hijacking, context window checks, high token usage alerts and questions held as secured configuration.
  • Repeatable deployment — The pipeline defined as infrastructure as code so it can be deployed reliably into the client's own AWS environment.
  • Handover — A working demonstration, materials and documentation delivered to the client, with their technical lead engaged throughout.

With the concept proven on their own decks, the client have a clear, secure blueprint for bringing automatic widget population into their product, on an AWS foundation ready to support continued development of the client's service.

AWS Stack

Amazon Textract

For extracting information from uploaded sales decks.

Amazon Bedrock

For the large language model that answers the client's questions.

AWS Lambda

For serverless orchestration of the extraction and answering pipeline.

Amazon SQS

For buffering document processing within Textract service limits.

Amazon S3

For secure upload, intermediate storage and results, within an Amazon VPC.

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.