James Bentley Email and Phone Number

- Developing data pipelines and improving existing ones, for example reducing dataset generation times by up to 99%, and fully automating a manual fortnightly SQL pipeline which feeds business critical products. - Leading development of new storage architecture to migrate away from point based storage and towards flexible index-vector hybrid storage of geospatial data - reducing storage costs and computation requirements. - Championing the use of cloud and modern geospatial data science tools such as Databricks, Carto, and H3. Assisting colleagues with integration where possible through knowledge share sessions and workshops. - Championing the use of LLMs to enhance existing products, for example prototyping adding chat functionality to customer reports. Additionally hosting internal knowledge shares and hackathons, and attending external customer engagement sessions as Landmark’s AI representative. - Generating new datasets for customer reports, such as using LiDAR data to calculate roof metrics at scale for solar panel suitability modelling, and using predictive temperature data to serve daily maximum temperatures and heatwaves from the 2020s to the 2080s at a resolution of 5km2. - Heavy involvement in cloud migration and modernising data storage approaches, leading migration of on-prem Oracle-stored data to Data Lake, and migrating SQL Data Warehouse to Data Lakehouse.

- Creating automated data analytics tools drawing from multiple large datasets for outlier detection, baseline generation, and ridership forecasting.- Developing fare evasion evaluation models based on a variety of input sources and ML techniques (Ranging from simple linear regression to development of neural networks for non-linear regression of complex data).- Analysing messy log data from station gate lines to gain a greater understanding of passenger flow, having reopened this project after earlier dismissal by senior developers as infeasible.- Geospatial analysis and prediction of passenger behaviours.- Carrying out proof of concept explorations of both novel and existing datasets to explore their viability to provide new metrics. - Presenting findings through a variety of mediums such as written formal reports and in person presentations to all levels of stakeholders, as well as making recommendations and high level conclusions based on these findings.

Application of Data Science techniques for a variety of pan-TfL projects, ranging from ticketing analysis and fare evasion detection to automated reporting and outlier detection of the impact of ridership from Covid-19.TfL Engineering VR and AR portfolio management. Responsibilities include strategy development, project scoping, internal and external engagement, story boarding, GDD writing, and C# development for familiarisation, training, and PoC purposes.Cost Engineering of various project including the Bakerloo Line Extension. Additional projects:- Development of handheld interactive DMI data feed readers for decoding and displaying output from SE903c bitstreams for live testing.- BLE development for the Wayfindr project. Using BLE to aid the movement of visually impaired people through stations.- Reverse Engineering obsolete circuits, analysing current and capital costs, and redesigning obsolete systems using VHDL.- Sitting on and assisting with the Digital Railway Working Group, Schemes Mental Health Group, and Year of Engineering Board.- Assisting in organising TfL involvement and support of the Cross-Government Prosperity Fund.STEM involvement:- Heavy involvement with STEM and recruitment teams promoting Apprenticeships across several events.- Travelling to schools to give talks promoting STEM carers and Apprenticeships.- Sitting on and working with the TfL Year of Engineering Steering Group.- Single-handedly mentoring two teams of year 12 students as part of the Engineering Education Scheme (EES), one team of which won best overall project.

Part of the Degree Apprenticeship programme at Jaguar Land Rover placed in the Cost Engineering department.