Causa - Causal Intelligence

Approach

Scientific rigour meets pragmatism

01

Understand the decision

Every engagement starts with the decision that needs to be made. The model serves the decision, not the other way around.

02

Encode domain knowledge

We work with your experts to translate institutional knowledge into model structure: causal maps and constraints that reflect reality.

03

Quantify uncertainty

Point estimates hide risk. Our models show you the full range of plausible outcomes, so you can weigh upside against downside before committing.

04

Deliver clarity

Results are communicated in terms your team can act on, not buried in technical jargon or hundred-page reports.

Case Studies

Results that matter

Pricing

Pricing strategy for a leading B2B SaaS

Built a causal pricing model to understand true willingness-to-pay across customer segments, informing tier restructuring and value-based pricing.

+15% ARPU

3 Segments

Causal Inference

Discounting strategy for an online retailer

Developed a causal model to isolate the true incremental impact of promotions, separating genuine lift from pull-forward and cannibalisation effects.

+20% Revenue

2.4x ROAS

Decision Science

Dynamic provisioning for a cloud platform

Built a Bayesian demand forecasting model to optimise infrastructure provisioning, balancing cost efficiency against strict uptime requirements.

-18% Costs

-12% CO₂

Services

What we do

Causal Inference

Move beyond correlation. We design analyses using DAGs and modern causal methods that identify what actually drives your outcomes, enabling interventions that work.

Bayesian Modelling

Principled uncertainty quantification that incorporates domain knowledge. Full posterior distributions, not point estimates. Honest answers about what the data can tell you.

Decision Science

End-to-end support from problem framing through to model deployment. We translate complex analysis into actionable insights for the people making decisions.

Technology

Meet Kairos

Kairos is our causal reasoning engine: a proprietary framework that lets us build, validate, and deploy causal models faster and more reliably than traditional approaches.

Causal mapping Define the cause-and-effect relationships in your business so assumptions are transparent and testable
Robustness testing Stress-test conclusions against hidden factors and alternative explanations before you act on them
What-if simulation Ask "what if?" questions and get grounded answers before committing to a decision
Honest uncertainty Full probability-based estimates with no false precision, so you know how confident you can be in every result

Engagement

Ways we work

01

Fixed-scope projects

A defined question, a clear deliverable. We scope the problem upfront, agree on outputs, and deliver a model or analysis you can act on. No open-ended retainers.

Timeline 4+ weeks Typical for Pricing, attribution, forecasting

02

Workshops & training

Hands-on sessions that upskill your team in Bayesian thinking, causal reasoning, and decision modelling. Tailored to your domain and your data.

Format 1-3 day sessions Typical for Data teams, leadership

03

Embedded partnership

For complex or ongoing problems, we work alongside your team as an extension, building capability in-house while delivering results.

Duration 3-12 months Typical for Strategy, R&D, product

Team

Who we are

Prof. Dan Franks

Co-founder

Professor of data science at University of York. Former lead researcher at Improbable.

Dr Jordan Hart

Co-founder

PhD in applied statistics from University of Exeter. Former ML engineer at digiLab.

More about us

Contact

Let's talk

Whether you have a specific modelling challenge or just want to explore how a more rigorous quantitative approach could help, we'd like to hear from you.

hello@causa.tech