Skip to content
PredictifyAnalytics

Practice areas

  • Predictive ForecastingDemand, revenue, risk and capacity models
  • Data Strategy & GovernanceOperating model and architecture roadmap
  • Analytics EngineeringPipelines, warehouse and semantic layer
  • Decision IntelligenceKPI trees and board-grade reporting
  • Machine Learning & AIModels in production, with monitoring
  • Pricing & RevenueElasticity, promotion and margin recovery

How we engage

  • Decision AuditTwo weeks, fixed fee, written findings
  • Build SprintEight to sixteen weeks, outcome-linked
  • Embedded PracticeSix to twelve months, retained
  • All servicesThe full practice in detail
  • Talk to a principalReply within one business day

Featured

Published work

Eight reproducible projects

Real analyses on 12 public datasets. Every figure computed, limitations stated in full.

Sectors

  • Retail & Consumer GoodsDemand volatility, assortment, promotion ROI
  • Financial Services & InsuranceCredit risk, claims leakage, fraud
  • Healthcare & Life SciencesPatient flow, capacity, adherence
  • Manufacturing & Supply ChainDowntime, yield, inventory position
  • Energy & UtilitiesLoad forecasting, asset health, balancing
  • Technology & SaaSChurn, expansion, unit economics

Featured

Published work

Eight reproducible projects

Real analyses on 12 public datasets. Every figure computed, limitations stated in full.

Projects

Latest thinking

  • The forecast is not the decisionDecision Science · 8 min
  • Why your data team keeps rebuilding the same pipelineAnalytics Engineering · 6 min
  • A CFO's guide to model riskGovernance · 10 min
  • All insightsEssays, notes and field research

Research & media

  • Reproducible projects8 projects, findings and limits published
  • The pipelineEight stages, the code behind each one
  • DeliverablesThe same work in six reporting tools

Featured

Essay

The forecast is not the decision

Accuracy past a threshold is worth nothing unless the decision rule changes. Working backwards from the decision instead.

About
Esc
Suggested
ServicesSix practices: forecasting, data strategy, analytics engineering, decision intelligence, machine learning and pricing. Three engagement formats, fixed scope.
Suggested
IndustriesRetail, financial services, healthcare, manufacturing, energy and technology — the forecasting problems each sector brings, and where the error has a price.
Suggested
ProjectsEight reproducible analyses on twelve public datasets — 6,219,116 rows, 51 findings and 50 stated limitations. Public data, documented method, every figure computed rather than typed.The pipelineEight stages from raw file to published result. Each names the code that performs it, the libraries it imports, and what went wrong there.
↑↓ to navigate↵ to open48 pages and sections indexed
Book a consultation
Services 01Industries 02Projects 03Insights 04About 05Contact 06Book a consultation
Home/Accessibility

Accessibility

What has been built in, what has not been tested, and how to tell us where it fails.

Last updated 8 August 2026.

What is implemented

Every page has a skip link, a single <main> landmark, a declared language and a logical heading order. Navigation, mega-menus, tabs and the homepage slideshow are all operable from the keyboard: tab lists take arrow keys, Home and End; the slideshow takes arrow keys and has a visible pause control; focus is never trapped.

Focus is always visible — outlines are styled rather than removed. The current page and the selected tab are announced through aria-current and aria-selected rather than colour alone.

Motion

If your system asks for reduced motion, we honour it. Scroll reveals resolve immediately, animated counters jump to their final value, charts draw in one frame instead of animating, and the homepage slideshow does not auto-advance.

Charts

Charts are drawn to a canvas, which is invisible to assistive technology on its own. Every one of the 128 charts on this site therefore carries a written alternative describing the shape and the values, exposed as an image label. Most project pages also publish the underlying figures as a real data table, which is the better artefact.

What has not been done

There has been no formal WCAG 2.2 audit and no testing with a screen reader, a screen magnifier or voice control. Colour contrast was designed for but has not been measured systematically across every state. The chart alternatives describe trends well and individual data points less well.

We would rather say that than claim a conformance level nobody has verified.

If something does not work

Email azlan@predictifyanalytics.com with the page and what happened. We will reply within one business day and tell you whether it is fixable quickly or not. If you need anything on this site in another format, ask and we will send it.

PredictifyAnalytics

A decision intelligence consultancy. We build the forecasts, models and reporting that let leadership teams commit to a number and defend it.

Services
  • Predictive forecasting
  • Data strategy
  • Analytics engineering
  • Decision intelligence
  • Machine learning
  • Pricing & revenue
Industries
  • Retail & consumer
  • Financial services
  • Healthcare
  • Manufacturing
  • Energy & utilities
  • Technology
The firm
  • About us
  • The pipeline
  • Contact
Research & media
  • Insights
  • Reproducible projects
  • Featured essay

© 2026 Predictify Analytics Inc. All rights reserved.

Privacy noticeTerms of engagementData ethics policyAccessibility