Itzel Soto

Leeds, UK · Open to new roles

Itzel Soto

Data Scientist & Engineer

Engineering-trained data scientist building data pipelines and AI-powered tools that solve operational problems and deliver measurable business impact. 12+ years of data experience across automotive R&D at Nissan and e-commerce analytics, now with an MSc in Data Science and Analytics from the University of Leeds.

Download CV Connect on LinkedIn Email me
Target roles
Data Scientist · Data Engineer · AI/ML
Track record
$2.4M cost reduction at Nissan
$14M+ attributed e-commerce revenue
Education
MSc Data Science & Analytics, University of Leeds
BSc Mechatronics Engineering, UDLAP
Languages
English (Excellent — full professional proficiency, IELTS C1)
Spanish (Native)

Featured projects

MSc dissertation · Industry partner CACI Ltd

Quantifying Street Attractiveness and Retail Performance

Found statistically measurable links between perceived street characteristics and retail performance across the UK

Problem
Is the look and feel of a street linked to how well its shops perform?
Approach
Built a UK data pipeline of 109,962 street-view locations with computer-vision-derived perception scores, aggregated them to road segments and spatially joined them to CACI’s retail data.
Tools & methods
  • Statistical modelling
  • Spatial joins
  • QGIS
  • Hugging Face
  • Mapillary & KartaView APIs
View details

Selected through the competitive GeoDS scheme to research this question with industry partner CACI Ltd (2025–2026).

The perception scores came from the existing Global Streetscapes dataset on Hugging Face, and I collected additional imagery through the Mapillary and KartaView APIs. My work was the pipeline, the aggregation to road-segment level, the spatial join to retail data, the statistical modelling and the mapping of results in QGIS.

Hackathon · University of Leeds, 2026

Pharmacy2U Data & AI Hackathon

Bronze Award, Challenge 2

Problem
Suggesting a patient’s likely next medication depends on their condition and prescribing history.
Approach
Co-developed a context-aware recommendation engine: a first-order Markov chain fitted to 3.8M prescription transitions, with condition-specific models.
Tools & methods
  • Markov chains
  • Recommendation systems
  • Streamlit
  • FDA RxNav API
View details

Built on the CMS DE-SynPUF synthetic claims dataset, with condition-specific models spanning 260,000 National Drug Code (NDC) nodes.

Delivered real-time inference through a Streamlit interface integrated with the FDA RxNav API, after team data preparation and exploratory analysis.

Industry · Email Marketing Strategist, MeMolaCreative

E-commerce Marketing Analytics

Contributed to $14M+ in attributed revenue across a portfolio of DTC brands

Problem
Direct-to-consumer brands need to know which email and SMS campaigns drive revenue, and how to improve them.
Approach
Analysed email and SMS performance across a portfolio of brands, using KPI trend analysis and comparative reporting to optimise campaigns and automated flows.
Tools & methods
  • KPI trend analysis
  • Comparative reporting
  • Campaign & flow optimisation
View details

A client engagement through Ecomcapitol, my e-commerce analytics consultancy. Tracked performance trends, compared results between campaigns and brands, and turned the findings into changes to campaigns and automated flows.

Experience

  1. 2025 – 2026 · Leeds, UK

    MSc Dissertation Researcher

    CACI Ltd, industry research project · University of Leeds

    Street-view data pipeline and spatial modelling of retail performance (see project above).

  2. 2018 – 2025 · Remote

    Founder & Analytics Consultant

    Ecomcapitol

    • Data-driven operations and performance analytics for 30+ direct-to-consumer brands.
    • Americanflat: 50% year-on-year DTC revenue growth (~$150K) through performance analysis and channel optimisation; managed 3 direct reports and 3 agencies.
    • Logico Cable: 20% year-on-year sales growth (~$60K) across five marketplaces and a Shopify store.

  3. 2011 – 2018 · Toluca, Mexico

    Engine Localisation Team Leader

    Nissan · previously Engine Management System Design Engineer and Engine Project Management Engineer

    • Delivered $2.4M in cost reduction through data-driven supplier analysis, quality-trend forecasting and ROI-based recommendations. Recognised with a Vice President Gold Award.
    • Led Bill of Materials costing and feasibility studies across 12+ new model projects.
    • Led international R&D assignments in Japan and Brazil.

Skills

Programming & data

  • Python
  • SQL
  • R
  • pandas
  • NumPy
  • Jupyter
  • Git & GitHub

Modelling & AI

  • Statistical & predictive modelling
  • Machine learning
  • Random Forest
  • XGBoost
  • TensorFlow
  • Deep learning
  • Computer vision
  • Hugging Face
  • LLMs & RAG
  • Recommendation systems

Engineering & data

  • Data pipelines
  • APIs
  • Process automation
  • Streamlit
  • HTML, CSS & JavaScript
  • AI-assisted development

Analytics & visualisation

  • A/B testing
  • Experimental design
  • KPI dashboards
  • Tableau
  • Looker Studio
  • Customer & marketing analytics

Geospatial

  • GeoPandas
  • QGIS
  • Spatial analysis
  • OpenStreetMap
  • Mapillary & KartaView APIs

Let’s talk

Hiring for data science, data engineering or AI/ML roles? I’d be glad to hear from you.