ProjectsHover a project to open it.

Problems solved,results delivered.

P / 01GTM operations · Inferenz · 2026
  • Problem. Selling AI solutions to US home health and hospice executives needed a repeatable outbound engine and clear visibility of what was working.
  • Approach. Built the sales stack (Apollo, HubSpot, Sales Navigator, JustCall, a custom Python tool), wrote a sequence sent under the CEO’s name, and designed the team’s daily outreach dashboard.
  • Result. About $500,000 in qualified pipeline from 2,380 emails, 1,200+ calls and 830 LinkedIn invites, with a 6.0% reply rate on the CEO sequence.
  • Value. Every rep and leadership now see which accounts to prioritise, which channel converts and each person’s activity at every stage.
ApolloHubSpotPythonSales NavigatorJustCall
P / 02Email deliverability · Inferenz · 2026
  • Problem. Outreach was landing in spam, so most prospects never saw it.
  • Approach. Diagnosed the cause, moved the team’s email from AWS to Microsoft 365 and standardised plain-text emails.
  • Result. Open rate rose from 7.5% to 74%.
  • Value. Outbound email became a top-five source of website traffic.
Microsoft 365Email deliverabilityGoogle Analytics
P / 03Marketing analytics · Ortem Technologies · 2025–26
  • Problem. A UK manufacturer couldn’t see which campaigns were wasting money.
  • Approach. Turned loose requirements into a marketing performance dashboard built in SQL and Tableau.
  • Result. Exposed about £1 million a year in wasted campaign spend.
  • Value. Gave the client a clear, evidence-based case for moving budget.
SQLTableauBusiness intelligence
P / 04Inventory forecasting · Ortem Technologies · 2025–26
  • Problem. Monthly inventory analysis took five people a full week.
  • Approach. Devised an inventory forecasting system in Python and Tableau, with demand forecasts built in.
  • Result. A one-click workflow that runs in one minute.
  • Value. The team can plan restocking ahead of demand instead of reacting to it.
PythonForecastingTableau
P / 05AI · Ortem Technologies · 2025–26
  • Problem. Thousands of customer reviews held useful signals, but reading them by hand was slow.
  • Approach. Launched an AI system that sorts reviews into product defects, delivery problems and customer preferences.
  • Result. Review processing time cut by 75%.
  • Value. Fed product fixes, delivery improvements and personalised recommendations.
LLMsPythonText classification
P / 06Data engineering · Contrado · 2022–25
  • Problem. Bringing a new data source or change into production took about a month and a half, and the data wasn’t always reliable.
  • Approach. Overhauled the pipelines in SQL and Python and standardised data quality with validation checks, duplicate removal and ERP-aligned naming.
  • Result. 3 days from new source to production, a 93% cut, with more accurate data.
  • Value. Fed 15+ live dashboards that replaced team-by-team Excel reports and returned 260 hours a year.
SQLPythonETLData quality
P / 07Customer analytics · Contrado · 2022–25
  • Problem. Marketing needed a sharper picture of who its customers were to target them well.
  • Approach. Directed customer segmentation with the marketing team using clustering and cohort analysis.
  • Result. Campaigns built on the new segments raised annual revenue by 10%.
  • Value. Turned customer-level analysis into marketing decisions.
PythonClusteringCohort analysis
P / 08Forecasting · Contrado · 2022–25
  • Problem. Leadership needed a dependable revenue baseline to plan against.
  • Approach. Built a time-series revenue forecasting model.
  • Result. Beat the original projections by 20%.
  • Value. A planning baseline the executive team could rely on.
PythonTime seriesForecasting