Industry Use Cases


Here’s a selection of real-world machine learning problems that we can help you to solve in a more open and customizable way. These examples are by no means exhaustive, so get in touch if you’re working on another use case.


Product Recommendations

Personalise the entire shopping experience to increase revenue and customer satisfaction.

  • Select from a suite of industry-leading recommendation algorithms including matrix factorization, content based, activity similarity, basket analysis.
  • Cross sell related products.
Next Best Action
  • Identify the next customer actions to maximise revenue.
Predict LTV (Lifetime Value)
  • Identify in real time customers who are likely to generate revenue.
  • Identify customers who are likely to be lost to prioritise proactive re-engagement.


Risk Analysis
  • Predict whether a customer will be low or high risk.
  • Supplement slow-moving credit scores with real-time insights based on customer behaviours.
Identify Emerging Trends
  • Identify new markets, trends or products from news and social media.
  • Provide early investment opportunities by spotting emerging trends.


Recommend insurance products
  • Predict which insurance products are appropriate for a new customer and upsell the most relevant policy.
  • Utilize social data and contextual data to classify users.
Fraud Detection
  • Identify fraudulent applications and commercial behaviour with pattern recognition and anomaly detection.
  • Reduce false alarms.


Content Recommendation

Show personalized up-to-the-minute content for each user to allow them to engage with the latest published articles or streaming media.

  • Optimize engagement and clicks through targeted recommendations for each user.
  • Select from a suite of industry-leading algorithms including cluster based, matrix factorization, content based, activity similarity.
  • Combine, A/B and optimize test sets of algorithms.
  • Integrate multiple recommendation areas per page.


Ad Personalization
  • Personalise the format of an advert – for example, customise the product, call to action and design.
  • Predicting affinity for brand or interest.
Ad Targeting
  • Deciding which audience to show which adverts
  • Personalised audience retargeting.
User Segmentation and Clustering
  • Gain actionable insights on your customers and audience by analysing the characteristics of the clusters.
  • Include segments defined by machine learning-based models into your marketing workflows.


Customer Support Triage
  • Classifying and prioritise customer support requests to reduce the overhead of manual triage roles.
Score Sales Leads
  • Build on your sales history to build a predictive model to identify the most profitable sales leads.
Real-time marketing
  • Optimize the customer journey to by orchestrating personalised multi-channel communication.

Move Faster with Seldon on Your Team

Our Enterprise solution works with you every step of the way, from analysis and requirements specification to scaling a machine learning production pipeline.

Let's talk
US: +1 (646) 397-9911
UK: +44 (20) 7193-6752

Find us

Rise London
41 Luke Street
view map

Join our newsletter

Receive updates, invites to events, and early access to product releases.