Joseph Osborne

Zapmap is a company dedicated to making EV charging easier, through providing their users with a map of 95% of all EV charge points, their current availability, and a method of payment all in one app. This time is focused on governance and resilience, as well as handling the implications of the UK PCPR which has overhauled the landscape for EV infrastructure.

• Scaled Analytics by embracing automated testing & close alignment with product function, building recurring revenue and a more trustworthy delivery culture.
• Built & Released first-of-its-kind utilisation product, providing daily insight for 100,000 EV chargers, set to support the UK market through its transition to a more open ecosystem.
• Implemented new Governance and delivery framework to ensure transparency & lineage while supporting agile delivery. 8x more changes per month, with a halved failure rate.
• Delivered new data agggregation platform for UK DfT enabling new generation of analytics and review of UK charging infrastructure using Zapmap data.
• Pioneered new integration approach to accelerate new partner onboarding by 10x to a 'Next Day Launch' pattern, unblocking growth from ~50 data partners to ~100 in a year.

This time is focused on scale, and coping with explosive growth, as the UK market accelerates, and as Zapmap launches to another 5 EU countries, the continental US and Canada.

• Hired, managed & mentored 5 other Data Managers with varying experience, forging them into a tight-knit team lauded internally for our commitment to results and support of each other. Finding & championing the right potential from within the business, and finding gems from the wider market.
• Use machine learning to accelerate data review processes, allowing small operations team to quality control 10,000+ locations per month
• Designed and built new data warehouse & delivery re-structure for Insights, leveraging BigQuery to accelerate outcomes. With ELT/ETL techniques and a focus on accessibility & scalability, positioned the data delivery ready to deliver internationally and prepared for a Real-time future.

This time was all about transformation & scaling to handle the imminent future, solving key issues in preparation for initial launches internationally to 5 EU countries.

• Transformed integration platform from simple scripts into scalable workflow in AWS, leveraging tools like SQS and DynamoDB to make scaling to millions of users simple. From 10,000 events per day to 750,000 events per day and growing.
• Built a new technical partnership with the developers, using small agile teams, close collaboration with specialists and clear priority frameworks. This model was later adopted across the business.
• Tackled delays in one of the key data product lines, bringing execution times down from 20+ hours to <10 mins and development cycles down from multiple months to days. Leveraging DAG (Directed Acyclic Graph) techniques and parallel execution to maximise efficiency.
• Transformed their Vehicles Data solution, designing and overseeing the implementation of a future-proofed and scalable data service

Kin + Carta: Edit is a company all about digital techniques to empower, and improve the customer experience. With an array of clients across many industries they specialise in building and maintaining great technical solutions, with skills hard to come by in house.

• Created brand new marketing segmentation using unsupervised learning techniques, widely adopted throughout a major client's whole strategy
• Took the data science team into a operating process which was version controlled and thoroughly documented, focusing on declaratively defined reporting and systems
• Completed a transformation in the team to a new set of technologies for reporting, simplifying overall process and saving 80+% of our licensing costs
• Automated swathes of formerly highly unreliable reporting to improve consistency, freeing up the team to deliver deeper more insightful work which produced millions in incremental revenue.

A small webserver which implements the Locations Module of the OCPI protocol, which can be used to host location data.

• Responsive: <100ms with 80,000 locations
• Low resource use: Runnable on Raspberry Pi, or $4 Digital Ocean Droplet

A commandline application which connects to a SQL information schema and extracts the information to the D2 diagramming format

• Supported refactoring process of legacy DB into cleanly seperated schemas
• Supported Access control audit for purposes of finalising a Microservices deployment

A web and local application which allows you to maintain a series of text blocks in a way which can be distributed neatly across multiple hosts

• Successfully distributes text blocks anonymously across locally networked machines, and across the internet
• Allows other agents to include new blocks in the chain, which the authors can select as the new principal