Engineering / Selected work

Case studies.

Advertising budgets, recruitment and product-reference search. A closer look at the business decisions and manual work we build AI systems around.

01Performance marketing

Decide where your ad budget goes.

In development

Which advertising partners bring valuable traffic, and where should the next budget increase go? We’re building quality scores and allocation models to help marketing teams compare partners and plan spend.

The approach

The work brings campaign data, traffic-quality signals and budget constraints into the same decision process. Teams can assess allocation options against performance and margin requirements.

Compare partners / Score traffic quality / Plan spend

02Recruitment agencies & talent teams

Build a shortlist from a pile of CVs.

Upload CVs and a job description. The platform compares candidates against the role’s requirements, explains the matches and drafts personalised outreach for a recruiter to review.

The approach

Recruiters can see how candidates score against the role’s criteria and read the reasoning behind each match. Candidate profiles, job requirements and outreach sit in one workflow.

Upload CVs / Review matches / Prepare outreach

03Brand protection / Research prototype

Support counterfeit checks with product evidence.

A customs officer has a photo of a suspected counterfeit. Our prototype finds the closest authorised products and their reference images, giving a reviewer evidence to compare before making a judgement.

The approach

Reviewers can inspect product images, packaging and tags from the reference catalogue. A close visual match identifies a relevant reference; the authenticity decision remains with the human reviewer.

Submit a photo / Find product references / Review evidence