available immediately · open to interesting work & collaboration

I help organisations
build analytics they can
actually act on.

Data scientist with 6+ years across government policy, AI safety, occupational health and HR analytics. I care about rigorous methodology, reproducible pipelines, and findings that don't get lost in a slide deck.

Let's talk See the work
benchmarking & driver modelling regulatory evaluation design AI risk analytics predictive systems data platforms & governance
analyst = { focus: "decisions that hold up", experience: "6+ years", location: "Manchester, UK", domains: ["HR analytics", "data platforms", "AI policy", "biometrics", "evaluation"], status: "available immediately" }
about

Rigour first.
Clarity always.

I specialise in turning complex, high-stakes data problems into analysis that decision-makers can trust. That means defensible methodology, not just impressive numbers — and findings communicated clearly enough that they actually change something.

I've worked across government policy evaluation, UKRI-backed research, occupational health data platforms, and large-scale organisational analytics. The thread across all of it: someone needed to make a decision, and the data wasn't yet in a shape that made that possible.

6+
years experience
£3.2M
govt. project scoped
4
industries worked across
MSc Data Science
Merit · 67%
University of Manchester, 2023
BTech Information Technology
Distinction · 92%
Anna University, 2019
experience

Where I've worked.

Oct 2025
Aug 2026
↑ most recent
IPRS Group
Data Scientist
Occupational health and workforce wellbeing group. Concurrent workstreams across IPRS Health and Robertson Cooper.
IPRS Health: Reporting and Data Platform
  • Saved 40 hours of team effort per week by migrating 255 legacy Excel reports onto a single Power BI framework, redefining KPIs with 8 team leads
  • Delivered a reusable data product covering 450,000+ patient records across eight teams, building a cloud data warehouse in Microsoft Fabric with Gen2 Dataflows and automated pipelines
  • Established a single source of truth across 90+ tables through star schema design in SQL, Power Query and DAX
  • Standardised auditable patient data handling organisation-wide through a layered governance framework of pseudonymisation, row-level security and role-based access control
  • Trained 6 analysts to a shared coding standard via SQL and Fabric bootcamps
Robertson Cooper: Client Workforce Analytics
  • Analysed 56,000+ records of employee wellbeing survey data with SQL and Python, benchmarking each client against Robertson Cooper's own benchmark dataset
  • Developed regression models that explained ~75% of the variation in productivity, wellbeing and staff turnover, helping clients identify the drivers to prioritise
  • Deployed the models and risk hotspot analysis across 12 health measures as a reusable Python pipeline that refreshed with each new client dataset, presenting to executives at eight client organisations
Power BIMicrosoft FabricSQLDAXPythonData Governance
Feb 2024
Sep 2025
Age Check Certification Scheme
Applied Data Scientist
(promoted from Data Scientist)
Independent certification body for age assurance and biometric technology. Worked with regulator, vendor and academic partners.
  • Improved biometric computer vision model accuracy from 61% to 81% by tracing errors to source data and fixing them through deduplication, labelling correction and validation rules in an AWS S3 pipeline
  • Led the data work on a £3.2M government-funded Australian evaluation of 52 biometric systems, designing sampling, validation and A/B testing across 2,500+ participants and analysing results with AWS Athena and Redshift
  • Contributed to scalable data science infrastructure (SQL and PySpark pipelines on AWS S3) for UKRI-funded deepfake detection research, curating 200,000+ image records across two jurisdictions under GDPR
  • Co-authored the published evaluation proposal report setting out the trial methodology, and presented findings to 50+ technical and non-technical stakeholders
PythonPySparkOpenCVAWSA/B TestingGDPR
Jan 2023
Jun 2023
Earthchain
Data Consultant (pro bono, alongside MSc)
  • Designed a weighted scoring model for 1,500+ carbon credit projects, scraped and validated from 3 ESG registries into MongoDB, to support investment decisions
  • Led six people to deliver 7 Power BI dashboards, the top-ranked outcome across all teams on the programme
Power BIPythonMongoDBWeb ScrapingAgile
Jun 2019
Aug 2022
Quinbay Technology
Senior Software Development Engineer in Test
  • Created UI and API test frameworks (pytest, JavaScript, Appium) that increased automation coverage by 80%
  • Streamlined releases with a Jenkins CI pipeline, saving 20+ hours per sprint
PythonpytestJavaScriptJenkinsCI/CDAppium
projects

Selected work.

Reporting data platform: sources into a Microsoft Fabric warehouse, star schema modelling, Power BI reporting, governed at every layer
Reporting Data Platform & Governance
↗ read more
Eight teams were running 255 separate Excel reports with no agreed definitions. I helped migrate them onto one Power BI framework built on a Microsoft Fabric warehouse, with star schema models across 90+ tables and governance applied at every layer, saving 40 hours of team effort a week.
Microsoft FabricPower BISQLDAXData Governance
Golden Dataset Pipeline: extract from a synthetic case management database, source checks with pandera, pseudonymisation, output checks and PII scan, deterministic train and eval split, versioned Parquet output with manifest
Golden Dataset Pipeline
↗ read more
A demo pipeline turning a manual dataset of completed legal matters into a versioned, pseudonymised, quality-checked and repeatable dataset, with evaluation data for measuring whether an AI model improves. Synthetic data only.
PythonSQLpanderaParquetGitHub Actions
Wellbeing Benchmark Dashboard
Wellbeing Benchmarking — Robertson Cooper
↗ read more
A people team needed to know which employee groups were at risk — and why. I built a benchmarking pipeline across 56,000+ survey responses, then developed predictive models explaining ~75% of variance in productivity and attrition.
PythonScikit-learnGit CI/CDBenchmarking
Australian Age Assurance Technology Trial
Australian Age Assurance Technology Trial
↗ read more
The Australian Government needed a defensible framework to evaluate 52 biometric vendors for Social Media age-restriction law. I designed the sampling methodology, built the CV pipeline, and co-authored the 88-page public evaluation proposal.
PythonSQLOpenCVAWSPolicy
Deepfake Detection Research
Deepfake Detection — UKRI Research
↗ read more
I built the data architecture, managed GDPR-compliant pipelines across two jurisdictions, and translated findings into practical risk mitigation — presenting to 150+ professionals across three workshops.
PySparkAWS S3GDPRTensorFlow
MSc Dissertation
Predicting Dashboard Usability — MSc Research
↗ read more
I ran controlled A/B experiments across 1.2M+ interaction logs and built a model to predict engagement quality before deployment — cutting UX friction by ~62% and giving teams a repeatable evaluation method.
SVRA/B TestingPandasUX Research
skills

My toolbox.

Python & ML

Pandas / NumPy
Scikit-learn
TensorFlow / OpenCV
PySpark

Data & Cloud

SQL & NoSQL
Microsoft Fabric
AWS (S3, SageMaker)
Azure & GCP

Analytics & Viz

Power BI & DAX
Experiment Design
Data Governance
Regulatory Eval.
project work

Focused engagements.

I occasionally take on scoped project work — particularly where there's a clear analytical problem and a team that wants something they can own and build on.

analysis
Analysis & Reporting
Turn your existing data into a structured report your team can use. Useful for survey results, performance data, or anything sitting in a spreadsheet waiting to mean something.
  • Structured analysis tailored to your question
  • Charts and visualisations included
  • Written report with key takeaways
  • One round of revisions
→ get in touch
build
Dashboard & Pipeline Setup
A dashboard your team checks weekly, or a pipeline that keeps reporting up to date. I'll build it, document it, and hand it over ready to use.
  • Power BI or Tableau dashboard
  • Automated data pipeline (Python / SQL)
  • Documentation & handover
  • Short-term support after delivery
→ get in touch
contact

Let's build something.

I'm always happy to talk about interesting problems — whether that's a potential collaboration, a project, or just swapping notes on something you're working on. Drop me a message and I'll come back within 24 hours.

✉ Say hello ◈ LinkedIn ⌥ GitHub ↓ Download CV