Work

Everything here is about the same problem: taking a commercial decision that gets made by intuition, modelling it, optimising it under real constraints, and putting it somewhere it can actually be used.

Applied AI & production governance

Case study

A decision platform for a DTC business

Built and run end to end by Anneal. Unified order, advertising, logistics and operational data in Postgres; a Bayesian marketing-mix model with channel adstock and saturation; an efficient-frontier spend optimiser; and a production AI agent operating under scoped permissions, identity controls, shadow-mode testing and user-acceptance gating.

Read the case study →
Method

Agent evaluation — the work nobody posts about

Why the distance between an AI demonstration and a production system is mostly evaluation, permissions and regression risk.

Read the piece →

Decision optimisation in regulated environments

External · FICO · 2019

Australian home loans: why analytic pricing is essential in 2019

Written by Doryan Gowty for FICO’s decisions blog while consulting to retail banks across APAC. Covers the ACCC’s Residential Mortgage Pricing Inquiry, discretionary pricing and transparency, price sensitivity across portfolio segments, and prescriptive analytics for rate-setting under regulatory constraint.

Read the article →
Method

Price optimisation in regulated environments

How price optimisation actually reaches a live bank decisioning environment: the problem shape, the governance and testing gates, and the failure modes that stop most programmes short of production.

Read the piece →