Maria Moya Email & Phone Number

Stamford, Ct, Us

Boston, Massachusetts, Us

* Presidents Club Leadership Award 2022* Managed an AI pod with 4 direct reports. This involves growth plans, 1:1s, promotions for my direct reports, hiring, driving ML strategy, working with legal on patents, working cross functionally with fullstack pods and e2e system design* Democratized AI at Drift to influence more than our human-in-the loop chatbots. Helped identify and obtain cross functional buy-in for product features that could use ML and worked cross functionally with full stack application pods to deliver these initiatives (i.e using AI for lead prioritization) * Regularly designed and led the construction of large scale infrastructure (i.e, real-time feature stores, scoring,etc.). These tier 1 launches led to a 200% increase in customer engagement within our product

Boston, Massachusetts, Us

Built EKS services, infrastructure for data pipelines(s3, RDS,dynamo,etc.) via terraform, enrichment via elasticsearch, NER e2e pipeline, training and deployment for Generative AI (this involved implementing a computationally efficient LED model and setting up/running distributed GPU training) as well as semantic search using sentence embeddings and ANN* Tech Stack: python, HuggingFace, PyTorch and sklearn*Orchestration: Kubeflow and Argo*CI/CD: Circleci and ArgoCD*Observability: Prometheus, Grafana, Logstash/Kibana*Offline Training: Kubeflow, Argo Workflows*Offline Data Enrichment: Argo Workflows, AWS Glue, EMR, RDS, Aurora, s3*Online Data Enrichment: SNS, SQS, Elasticsearch, Dynamo*Model Serving: Kserve

Woonsocket, Ri, Us

Currently working on building data pipelines, Machine Learning and Deep Learning models at scale for nearly 90 million active customers and 2 years (or roughly 12 TB) worth of transactional data by customer I predominately use spark(pyspark) and python for data preprocessing as well as Keras(TensorFlow backend) and MLlib for modeling In addition to modeling, another key aspect of my role involves collaborating with DevOps and engineering as far as productionization of our models

Boston, Ma, Us

Contract/Temp roleArtificial Intelligence TeamBuilt an AI algorithm that was able to detect numerical and text based data inconsistencies among 2000+ daily published pdf reports

Miami, Florida, Us

• Python, linux and AWS based environment. Normal day-to-day involved working with large GPS mobile data, scraping data and building machine learning analytics as well as data visualization tools via Python• built an interactive tool that allowed clients to look at high versus low potential business areas, specifically areas in which they should place their new store. This project included: scrapping google lat lon points, clustering them into business areas, feature engineering the business area shapes and computing a site potential ranking/score• Provided machine learning analytic tools such as: locations in which our clients could seek new potential customers, where a client should place their next store, how quantitatively measure what political interest people care about by city in the US,etc.

Boston, Ma, Us

•Single-handedly built a production ready machine learning algorithm (via Python) that was able to predict around 72% of all uninsured claims. This algorithm saved us around $175,00 per year• Led all machine learning initiatives and presented any insightful findings to the VP and other claim directors absent quantitative jargon• Matained/debugged our T-SQL stored procedures and SSIS packages. Led a large scale data migration project to include an entirely new book of business• Constructed Tableau dashboards/interactive data visualization tools for stakeholders

Medford, Massachusetts, Us

• Taught two sections each consisting of approximately 30 students• Led and prepared discussions on introductory microeconomic and macroeconomictopics

• Pioneered the population studies project, which involved visually representing our timeseries tabular data into animated spatial maps via ArcGIS• Wrote Python scripts to automate processes via ArcGIS. As a result our map generatingprocess was around 40% faster• Trained and managed new employees. Helped grow a team of 8 new employees• Wrote VBA macros in order to update/format 7,000+ datasets instantaneously andmanage our database more efficiently. This reduced our reporting time by nearly 83%

Master Of Science (M.S.), Econometrics And Quantitative Economics

Bachelor Of Arts (B.A.), Economics And Mathematics