Maryam Pouryazdan
Data Science Manager | AI & Smart Technology
San Francisco,CA
Summary
Innovative Data & Smart Technologies Leader with 6+ years of experience applying machine learning, AI, and IoT solutions across utility, manufacturing, and customer analytics domains. Proven track record in deploying predictive models, digital infrastructure, and real-time monitoring systems that deliver measurable impact. Skilled in Python, SQL, cloud platforms (Azure, AWS), and big data tools, with a Ph.D. in Electrical and Computer Engineering.
Overview
7
7
years of professional experience
2019
2019
years of post-secondary education
Work History
Data and Smart Technologies Manager
Veolia North America
10.2023 - Current
- Led cross-functional digital and engineering teams to design and implement AI-driven and IoT-enabled smart infrastructure solutions across large-scale utility and municipal systems, delivering measurable impact and innovation.
- Built predictive maintenance and time-series models leveraging SCADA, AMI, laboratory, and sensor data to proactively identify system anomalies—resulting in an 18% reduction in chemical usage and over 25% decrease in odor incidents across water and wastewater treatment facilities.
- Led the deployment of smart leak detection technologies using transient pressure and acoustic monitoring, enabling continuous surveillance and AI-powered analytics—achieving a 21% reduction in Non-Revenue Water (NRW) and enhancing capital planning through data-informed infrastructure investment decisions.
- Integrated Microsoft Azure, AWS, Microsoft Fabric, Power BI, and GIS into real-time dashboards and digital twin platforms, supporting operational visualization, monitoring, and proactive decision-making.
- Managed live deployments of smart water systems in multiple municipalities, driving process retrospectives and delivery enhancements to foster continuous improvement in mission-critical, high-reliability settings.
- Acted as a client-facing lead, translating complex data models and insights into accessible narratives and strategic recommendations through reports, dashboards, and executive presentations.
Senior Data Scientist
CSpace
04.2022 - 08.2023
- Conducted extensive A/B tests and multivariate tests for over 250 clients to optimize surveys' design and customer review campaigns with goal of higher Gen-Zs participation.
- Built a sentiment analysis tool using tfidf transformer to predict fraudulent customer behaviors on a large dataset of over 20k users and rate them which resulted in 15% save on payments to malicious users.
- Contributed towards moving the direction of reach market program from intuition-based to data-driven approach by partnership with client, customer acquisition and engineering team.
- Created a scalable segmentation framework using clustering methods in order to discover new segments of customers for more than 200 clients, resulted in increasing customers' loyalty and engagement with more than 85% accuracy.
- Partnered with business leaders across the organization to identify and unlock new sources of value from data and AI methods, including raising data literacy by democratizing access to curated datasets and ML self-service tools.
- Developed a scoring model using Random forest to select high value customer from our community members resulting in recruiting up to 25% more influential customers.
IoT Data Scientist
Watts Water Technology
10.2018 - 03.2022
- Identified potential customer by deploying a mapping and forecasting system by the use of K-Means clustering and Gradient Boosting Classifier algorithms which led to target relevant customer with AUC-ROC score of 74%.
- Consults with business partners to understand problems and goals, understand business value, and translate into business opportunities to be captured through predictive analytics solutions.
- Used matrix profile-package to detect time series motifs and anomaly behavior of IoT devices and identified overall health of devices to predict regular maintenance which saved up to 18% on repair costs.
- Implemented various time series forecasting model to predict market demand which surfaced the key data points such as seasonal fluctuations, consumer trends, commercial construction market, weather-data and purchase correlations, led to save up to $255k yearly.
- Applied MARS regression model to predict remaining useful life of IoT devices which improved service levels, revenue and customer experience by 32%.
Education
Ph.D. - Electrical And Computer Engineering
Clarkson University
Skills
ML Models: Time-Series, Regression/Classification Models, NLP, Reinforcement Learning, Neural Network, Clustering, Sentiment Analysis
Optimization algorithms: genetic algorithms, reinforcement learning
Programming and Script: Python(pandas, scikit-learn, NLTK/spaCy), SQL, R, Scala, Java, LinuxShell Script, Matlab, Tableau
Statistical Skills: A/B Testing, Hypothesis Testing, Sig Testing, Driver Analysis, Confidence Interval, Price Sensitivity Analysis
Big Data Ecosystem: Microsoft Azure, Apache Spark (Spark Segmentation, Spark SQL), Amazon Web Services (EMR, EC2, S3)
Timeline
Data and Smart Technologies Manager
Veolia North America
10.2023 - Current
Senior Data Scientist
CSpace
04.2022 - 08.2023
IoT Data Scientist
Watts Water Technology
10.2018 - 03.2022
Ph.D. - Electrical And Computer Engineering
Clarkson University
Similar Profiles
Jack BaldersonJack Balderson
Environmental Services Technician III at Veolia North AmericaEnvironmental Services Technician III at Veolia North America
Jonathan DeSimoneJonathan DeSimone
Engineering Intern at Veolia North AmericaEngineering Intern at Veolia North America
Ross GambinoRoss Gambino
Deputy Team Leader at Veolia North AmericaDeputy Team Leader at Veolia North America
Gerardo PenaGerardo Pena
Lead Operator at Veolia North AmericaLead Operator at Veolia North America
Perla EscobarPerla Escobar
Cleaner at J&r ConCleaner at J&r Con