Case Studies | SaaS B2B

Personalised, multilingual video at scale on AWS

Personalised Recommendations and Multilingual Subtitles

About

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

Challenge

The challenge had three focus areas, all centred on scaling a personal, accessible experience without scaling the manual effort behind it.

Relevant recommendations per user

Recommendations needed to reflect how each person actually engages, watch time, likes, comments and shares, and update as that behaviour changes, not sit on a static model.

Accessibility across languages

Every the client needed captions in its source language and translations into others, so content is understandable to the whole audience. Doing that manually per video simply does not scale.

Discoverability with low overhead

Users needed to find the right the client across a growing, multilingual library, and the whole thing had to run automatically, securely and cost-efficiently.

Solution

When a creator uploads the client, it lands in Amazon S3 and triggers an AWS Lambda that uses AWS Transcribe to detect the source language and generate. VTT captions, then runs an Amazon Bedrock batch job to translate those captions into a set of target languages, storing every version alongside the original for on-demand playback. The same upload is stripped for metadata and fed into Amazon Personalize.

Per-user

Recommendations from live engagement signals

Auto.VTT

Captions generated on every upload

Multilingual

Batch translation for on-demand playback

By the numbers:

  • Per-user - Recommendations from live engagement signals
  • Auto. VTT - Captions generated on every upload
  • Multilingual - Batch translation for on-demand playback
Changes

The proof of concept delivered an end-to-end serverless platform that met its success criteria: per-user recommendations from live signals, a fully automated caption-and-translate pipeline, and semantic search across a multilingual library, all deployed as infrastructure as code.

  • Personalised in real time — Amazon Personalize turns live signals, watch time, likes, comments and shares, into per-user recommendations.
  • Captioned automatically — Every new the client is transcribed to.VTT in its source language via AWS Transcribe, with no manual step.
  • Multilingual by default — Amazon Bedrock batch-translates captions into target languages, stored alongside the original for on-demand selection.
  • Searchable by meaning — Translated subtitles are indexed in a S3 Vectors store, so users find relevant Slicks across languages.
  • Serverless and resilient — Event-driven Lambda with idempotency, retries and least-privilege IAM, deployed via CloudFormation.
  • Handover — Delivered with documentation and Loom walkthroughs, ready to deploy into the client's own AWS account.

With the platform proven, the client can roll it out across their growing library, adding languages and tuning recommendations as the audience scales, without adding operational overhead.

AWS Stack

Amazon Personalize

For per-user recommendations driven by live engagement signals.

AWS Transcribe

For automatic.VTT caption generation in the source language.

Amazon Bedrock

For batch translation of captions into multiple languages.

AWS Lambda

For event-driven orchestration, with Function URLs for ingestion and retrieval.

AWS CloudFormation

For automated, repeatable deployment of the whole stack.

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.