Evan Blackie
Principal Data Scientist @NZ Transport Agency Waka Kotahi
Signup · Get unlimited contacts
WORK HISTORY
Principal Data Scientist @NZ Transport Agency Waka Kotahi
Wellington, NZ
Contract role within Safety Camera System implementation programme (Evaluate/Data Science and Modelling):• Developed code to manage camera infringement volumes demand against operational resourcing supply (Dynamic Infringement Management System, DIMS), including data pipeline, randomisation algorithm and implementation within business process.• Building analytics pipelines with Azure Databricks/Spark, SQL, Python/R for reproducible analytics combining data from safety camera systems, crash data (CAS), external data vendors via API (Tomtom), GIS road networks, adhoc data.• Time series prediction of speeding offences across safety camera sites using traditional (ARIMA, lag models) and neural network (multilayer perceptron, LSTM) prediction methods using R and Python (Keras/tensor flow).• Validation of rubber tubes speed data compared with ‘ground-truth’ camera data using road-side tests at vehicle-testing track to understand measurement error in tubes data (hierarchical mixed model).• Developed Python package to automate collection of vehicle speed data from Tomtom API (a source of speed and travel time data through commercial vendor) across multiple physical locations for preset dates/times.• Created data specifications for benefits realisation framework to show performance of safety camera system following operation of new camera sites (data definitions, database derivation etc).• Implemented automated curve-fitting algorithms to convert TomTom percentiles data (an aggregate format) into an inferred speed distribution for visualisation and estimation of speeding compliance and offences.• Data discovery, validation, testing and documentation across multiple new data sources (raw formats and post-ingestion, creation of interim data model (star-schema modelling) and testing (functional, integration, UAT).• Implemented PowerBI dashboard to monitor baseline speed data for average speed cameras.
EDUCATION
University of Otago
BSc (Hons), Chemistry
Victoria University of Wellington
PhD, Physical Chemistry
SKILLS
ABOUT EVAN BLACKIE
Overview -15-years+ experience in data science using statistical and analytical techniques in research, government and business environments- Pursuing roles in data science, machine learning, AI with an applied focus- Have a strong appetite for detail and complexity, love coding, modelling and problem solving generally- Attuned to business value and forming productive relationships across key stakeholders- Versed in Agile/Devops/SDLC and team practices- Key Skills - Applied Data Science - Experienced across the full model lifecycle (experiment design, feature engineering, training, validation, deployment and monitoring) using traditional statistical methods and machine learning- NLP/Text Analytics -Practical NLP/semantic analysis using topic modelling via Latent Dirichlet Allocation, including tokenisation/stemming, named‑entity recognition, and via embedding‑based approaches (e.g. BERT, RoBERTa) for semantic similarity/clustering. Used for summarisation/classification, search‑and‑match, and insight extraction on publicly available data- Evaluation -Formal statistical and evaluative methods for estimation, prediction and causal inference including regression techniques, propensity score matching, randomised controlled trials (A-B testing), quasi-experimental design/natural experiments- Data Pipelines -Creation of data pipelines across multiple environments (Azure Databricks, Snowflake/dbt, SSMS). Familiar with governed enterprise envrionments (Azure, Snowflake, SQL Server) and typical Data Modelling practices (star-schema, data vault, normalised/denormalised). Understanding of SDLC, containerisation, environment reproducibility, automation, devops/orchestration for delivery- Business Experience - Experienced in using analytics for actionable insights through insight packs, predictive/descriptive models and custom visuals/dashboards. Comfortable leading analytical workstreams, mentoring teams, and working across technical, policy, engaging operational stakeholders, including ethics/governance processes- Programming and Platforms -Languages: Python, R, Azure Databricks/Spark, SAS (Viya, EG), SQL (SSMS/RS, Snowflake, Oracle), dbt, SPlus, Matlab. Text analytics, NLP, Keras/Tensor Flow, network analysis. Agile/productivity: Gitlab/Github/Azure Devops, JIRA/Confluence slack, docker/containerisation/VMs. BI tools:(PowerBI dashboarding, Sharepoint, R Shiny, Meridio, Crystal Reports, Tableau and GIS software (QGIS, ArcGis).
This profile is compiled from publicly available professional sources. Unifers is not affiliated with or endorsed by LinkedIn. Request removal of this profile.