I have five years of experience using data science techniques to help business strategy and product development across multiple industries, including crypto, marketing, and insurance. My area of expertise involves using novel techniques in data analysis, machine learning, statistical inference, and causal inference to identify and address complex business opportunities, problems, and areas of potential growth.As a manager of data science teams at fast-growing startups, I have collaborated on various aspects of the business strategy, including new product development, anti-fraud initiatives, customer support efficiency, customer lifecycle management, and marketing. In all of these, my goal has been to create models that deliver tangible value to the end-users. To achieve this, I have followed a data strategy framework that involves several steps:1. Problem identification and transformation- Collaborate with stakeholders to determine how to address the problem using the available data.- Translate abstract business challenges into a series of well-defined data science projects.2. Data quality assurance- Work alongside data engineers to ensure that the data used for analysis is of high quality and relevant to the problem at hand.- Identify the data required for building future models.- Conduct exploratory data analysis3. Machine learning model development- Apply novel machine learning techniques to build robust models that can solve the identified problem.- Use appropriate metrics to measure the model's performance and ensure that it meets the intended scope.- Create autonomous and flexible models that are able to future versions.4. Machine learning deployment- Collaborate with MLOps engineers to build the infrastructure required to deploy the models.- Establish triggers to detect different types of model drift and enable automatic retraining when possible.5. Impact evaluation- Design a series of experiments or feature launches to assess the model's impact on business outcomes.- Establish a framework for sharing the results with the business intelligence team.In summary, my experience in leveraging data, machine learning techniques, and statistical inference to solve complex business problems, combined with my ability to collaborate across different parts of the organisation, have enabled me to manage the creation of models that deliver value to the end-users.I am always happy to talk about data science at manucortes95@gmail.com or here www.linkedin.com/in/manuel-cortes-barrios
-
Head Of Machine LearningBitso May 2023 - PresentMéxico, Df, Mx -
Senior Data Science Manager - ProductBitso Sep 2022 - May 2023México, Df, MxLed the full lifecycle of machine learning initiatives that aimed to improve the current product strategy or create new features to enhance users’ experience. • Headed the development of an investment product that helps users discover better diversified investment strategies, resulting in an 83% increase in the risk-reward trade-off against holding BTC.• Led the end-to-end development of a new Profit and Loss feature, culminating in a successful launch with a 15% revenue increase and a 30% increase in recurrent app usage compared to the control group.• Coordinated the creation of a VIP model for high value users, empowering customer support to prioritise tickets based on users' potential future value, resulting in a 25% increase in CSAT.• Championed the antifraud initiatives, overseeing the data requirements ultimately deploying a robust machine learning model in production, resulting in a significant 60% reduction in customer support tickets related to fraud.• Constructed the individual Lifetime Value model with an average RMSE of $1, facilitating effective segmentation and enabling a data-driven company's strategy.• Designed a robust causal machine learning model that demonstrated that recommending a "lower fees" product would not result in increased trading volume from users, prompting a decision to halt the initiative before launch.• Developed a comprehensive top-line revenue model that illuminated the revenue retention issue the company was grappling with, prompting a swift response from the stakeholders to prioritise and address the matter. -
Senior Data Scientist Ii - ProductBitso Nov 2020 - Sep 2022México, Df, MxDeveloped and put into production machine learning models that helped improve the business strategy.• Constructed the individual Lifetime Value model with an average RMSE of $1, facilitating effective segmentation and enabling a data-driven company's strategy.• Forecasted the revenue and retention probability for the upcoming month with an 85% accuracy rate, providing insights for a new product related to fixed income investment..• Predicted the likelihood of customer retention over the next months with a 93% AUC score, allowing targeted retention efforts for at-risk customers.• Built a customer segmentation framework to assess the likelihood of a user's ongoing profitability and the probability of returning after a cash-out with a 90% confidence. • Executed an experiment to identify the most effective retention strategy, resulting in a 10% reduction in churn probability.• Constructed a predictive model that achieved an 80% AUC score for determining whether a user had inadvertently opted out of receiving notifications due to a settings bug, allowing for targeted campaigns in conjunction with CLM efforts.• Developed an unsupervised time series model capable of classifying market conditions into three distinct states, enabling marketing teams to optimise campaign launches based on the prevailing state.• Enhanced the revenue model for users by leveraging Spark ML, resulting in a 40% reduction in RMSE and a significant decrease in processing time. -
Head Of Data ScienceKanto Studio Jul 2019 - Nov 2020Headed the development of the machine learning infrastructure needed for the creation of an AI marketing app, from data ingestion to final deployment into production of models.• Designed a machine learning model that classified marketing campaigns into distinct clusters, leading to an increase of 15% on average ROI for new launched adds.• Employed time series methods to develop a predictive model for ongoing campaign costs, resulting in a 30% reduction in average campaign expenses.• Devised an onboarding transition model, enabling promotions based on user progression probabilities, leading to a 40% increase in successful onboarding completion.• Collaborated with software engineers to improve the data pipeline and create new ETLs for ingesting marketing data from diverse sources, achieving a reduction in 60% the processing time.
-
Risk Consultant - Data AnalystSespec Aug 2018 - Jul 2019Federal District, MxEmployed both traditional risk models and cutting-edge machine learning techniques to improve prediction accuracy and automate manual processes. • Developed a priority index for freeway with 5 different predicted categories, resulting in a budget optimization for those clusters with the best cost-expected benefit trade-off.• Generated time series forecasts for car accidents, resulting in a 20% reduction in unnecessary reserves with a 95% level of confidence.• Devised a system with an 80% AUC to facilitate fast risk recommendations from engineers to clients based on infrastructure issues.
Manuel Cortes Education Details
-
Instituto Tecnológico Autónomo De MéxicoRisk Management -
Instituto Tecnológico Autónomo De MéxicoApplied Mathematics -
Instituto Tecnológico Autónomo De MéxicoActuarial Science
Frequently Asked Questions about Manuel Cortes
What company does Manuel Cortes work for?
Manuel Cortes works for Bitso
What is Manuel Cortes's role at the current company?
Manuel Cortes's current role is Head of Machine Learning | M.S. Risk Management.
What schools did Manuel Cortes attend?
Manuel Cortes attended Instituto Tecnológico Autónomo De México, Instituto Tecnológico Autónomo De México, Instituto Tecnológico Autónomo De México.
Free Chrome Extension
Find emails, phones & company data instantly
Aero Online
Your AI prospecting assistant
Select data to include:
0 records × $0.02 per record
Download 750 million emails and 100 million phone numbers
Access emails and phone numbers of over 750 million business users. Instantly download verified profiles using 20+ filters, including location, job title, company, function, and industry.
Start your free trial