Ivan Alvarez Email & Phone Number

London, England, United Kingdom

London, England, United Kingdom

London, England, United Kingdom

• Led a team of 4 data scientists (2 mid, 2 senior) as direct reports, building machine learning (ML) automation tools for BBC News content authoring, used daily by approximately 3,000 journalists and editors.
• Oversaw the development of automated metadata suggestion tools, integrated into existing content production pipelines, and empowering editorial decisions on where and how individual News stories are published.
• Managed cross-functional stakeholder relationships, including engineering, product, delivery, editorial and analytics functions, aligning data science work with product priorities.
• Championed data science across the business, emphasising the value added by data science applications to automating content generation, curation, and experience personalisation for end users.

London, England, United Kingdom

• Built recommendation engines to enable personalisation at scale for millions of users across news, sport and video products.
• Designed and productionised a hybrid algorithmic recommender for the BBC iPlayer video-on-demand product, showing 30% uplift over manually-curated recommendations in A/B test.
• New recommender provided improved personalisation for 11m weekly users and reduced editorial burden by 20h per week.
• In a survey of 500 representative users, 94% rated the quality of recommendations as "good" or "very good".
• Developed cross-product engine to providing real-time personalised recommendations between BBC News and BBC iPlayer products.

• Led a team of 8 researchers (1 graduate student, 2 staff scientists, 5 research assistants) developing supervised machine learning models for decoding physiological timeseries data with high prediction accuracy (>80%).
• Developed advanced computer vision models including spatio-temporal Gabor wavelet pyramids and convolutional neural networks in TensorFlow and sciki-learn for feature engineering across thousands of naturalistic images.
• Designed a regularised linear regression and permutation testing framework to fit encoding models that predict future brain responses in an extremely large data regime, with thousands of features and millions of.
• Optimised timeseries analysis software for deployment in a high performance computing (HPC) environment, leveraging GPU capacity for x20 speed-up over conventional implementations.
• Applied dimensionality reduction techniques (PCA, ICA, clustering) and multivariate analysis to explain how visual information is encoded in the brain.

Oxford, United Kingdom

• Spearheaded a 5‐year project using biologically‐plausible neural networks and brain imaging to understand the neural mechanisms of 3D vision & inform the design of VR kits.
• Studied neural mechanisms of binocular stereopsis in humans, using computational modelling, ultra-high field MRI, and behavioural psychophysics.
• Created a neural network model of 3D depth detection, which can accurately predict brain responses.
• Leveraged Gaussian mixture models, hierarchical clustering, and multi-class classification to predict the depth perception abilities of individuals based on their brain scan data, leading to successful outcomes for a.

London, England, United Kingdom

* Designed and validated a novel forward model for timeseries analysis of physiological data, now used by 7 teams worldwide and leading to 16 peer‑reviewed publications.* Developed algorithms for method fusion in MRI brain imaging.* Experience with clinical and pre-clinical data collection, analysis and interpretation.* Triaged requests for ad hoc data.

Doctor Of Philosophy - Phd, Computational Neuroscience

Master Of Science - Ms, Cognitive Neuroscience, Distinction

Bachelor Of Science - Bs, Psychology, 1St Class