Jan 2021 – Jan 2022 · 1 yr

• AI/Data Science/Coding:
• Analyzed Large Data using PySpark on Azure databases.
• Developed the following Models for Detecting Financial Crime:
• Computer Vision and NLP models to automate document extraction and
• classification, reducing manual review.
• Transfer Learning with VGG16 and BERT for document classification and
• information enhancement.
• Tax Evasion classification model using diverse data sources.
• Leadership/Others:
• Led and coached teams of data scientists in multiple departments.
• Investigated Deepfake using GANs and presented findings.
• Automated KYC processes with robotic process automation (RPA).

Sep 2019 – Dec 2020 · 1 yr 3 mos

Jan 2018 – Jul 2019 · 1 yr 6 mos · Rotterdam Area, Netherlands

• AI/Data Science/Coding:
• Analyzed Large Data in Hadoop environment on Azure machines.
• Predicted Smart Meter models, 2 million households, saved €1.7 million
• Clustered 2,000 Mechanics on driving behavior using GPS Sensor data.
• Forecasted Excavation Damage Risk using ML and Geographic programming.
• Developed and Deployed models for energy theft in Cannibal farms.
• Developed call center demand models for optimal resource planning.
• Leadership/Others:
• Led and coached an 8-member data science team at the CDO office.
• Provided strategic advice to CDO, COO, and CTO, defining use cases.
• Speaker/Organazier workshops for Directors and Lean Six Sigma experts.

Sep 2016 – Dec 2017 · 1 yr 3 mos · Duiven

• AI/Data Science/Coding:
• Analyzed Large Data from SAP Hana using Oracle SQL Developer.
• Developed Forecasting Models for Dutch Gas Network.
• Developed NLP Models for Energy Customers (Solar Panels etc.)
• Optimized Gas and Heat networks with 3D visualization.
• Constructed Gas Network Topology using all assets.
• Designed a Geographic User Interface for forecasting models

Apr 2014 – Sep 2016 · 2 yrs 5 mos

Mar 2009 – Mar 2014 · 5 yrs

• AI/Data Science/Coding:
• Analyzed Large Data using Oracle QL and TeraData.
• Developed the following models:
• Forecasting model for 30,000+ E-Commerce products, increasing accuracy
• threefold and reducing resource requirements.
• Assortment Differentiation
• Real estate Optimal locations incl. three years' sales estimation.
• Individual Customer Next purchases based on historical data.
• Customer segmentation, recommendation, and personalized offers
• Weekly Promotion items for three companies (AH, Etos, Gall&Gall).
• Lost sales and write-offs for 30,000+ products in 1000 AH stores.
• Nonfood demand forecasting for Albert Heijn XL stores.
• Conducted predictions and trend analyses for product groups.
• Optimize pricing and Assortment.
• E-commerce Profitability through customer clustering/targeting.
• Analyzed slow-moving products using time series models.
• Logistic supply and Optimal delivery times for long-deliveries
• Leadership/Others:
• Led and Coached 3 teams, with an average of 5 data scientists per team.

Jun 2012 – Sep 2016 · 4 yrs 3 mos · Utrecht Area, Netherlands

• AI/Data Science/Coding:
• Managed very large datasets from AWS, SAS/Teradata, and Hadoop cloud.
• Developed the following models:
• Forecasting models for tax correction for all Dutch taxpayers
• (around 11 million) in various tax sections.
• Tax fraud detection in collaboration with law enforcement agencies.
• Sentiment Analysis NLP models using data from Coosto, Twitter,
• and tax declaration communications.
• Clustering models for taxpayer behavior analysis and demographic
• profiling.
• Problematic Demographic Groups
• Leadership/Other:
• Led and Coached team of 4 data scientists in National Supervision dep.
• Defined and fine-tuned multiple use cases
• Designed interactive dashboards using Tableau and Power BI

Apr 2012 – May 2012 · 1 mo · Hilversum

• Core Business:
• Bus network company
• Activities:
• Pre-analysis of public transport data for optimization of bus routes
• Clustering of public transport customers in different demographic area’s
• Software skils:
• Matlab

Dec 2011 – Dec 2011 · 0 mo · Amsterdam Area, Netherlands

• Core Business:
• Media company
• Activities:
• Data analysis of data originated from Sanoma Uitgeverijen, Libelle,
• Libelle Academy
• Creating Online dashboard
• Software skills:
• Tableau Server, MySQL.

Feb 2011 – Mar 2011 · 1 mo · Utrecht Area, Netherlands

• Core Business:
• Europe payment service provider
• Activities
• Data analysis and modeling of Equens credit card/debit card payments
• The determination of optimal Threshold for Refund problem of Credit cards for ING
• The determination of optimal Threshold for Refund problem of Debit cards for ING
• Writing a scientific Test report
• Software skills:
• MATLAB, PL/SQL and Microsoft Excel.

Jan 2010 – Present · 16 yrs 4 mos · Amsterdam Area, Netherlands

• I am a PhD AI Scientist and Engineer with 20+ years of experience specializing in building agentic AI systems.
• My goal is to teach Data Scientists and AI Engineers how to build real-world AI agents that actually work through hands-on projects across diverse industries.
• After shipping 400+ production agents, I’ve refined a method that bridges the gap between AI theory and industrial reality.
• Watch my Free AI Agents Masterclass (30 Min) + get my 56-page Guide below: 👇
• https://www.maryammiradi.com/free-ai-agents-training
• Join my AI Agents Course Training 5-in-1, Project-based & Build real-world AI agents. 👉 https://www.maryammiradi.com/ai-agents-mastery
• Created by a real-world AI expert , PhD, 20+ years, Global Instructor in 90+ countries.
• I am the founder and Chief AI Scientist of Profound Analytics.
• BUSINESS SECTORS
• ▪ Banking | DFC | KYC | AML
• ▪ Energy (network)
• ▪ Retail | E-Commerce
• ▪ Asphalt, Road and Transport
• ▪ Government (Fiscal)
• ▪ Public transport
• ▪ Publishing houses
• SKILLS
• Recent Programming / Platform Tools:
• ▪ Python (SKLearn, Tensorflow, Keras,
• Pandas, Numbpy, Hyperopt, MLFlow,
• Opencv, Pypdf, passporteye, Spacy, NLTK,
• Bertje, NetworkX, GraphFrames,
• KarateClub, )
• ▪ PySpark | Databricks | AWS
• Sagemaker Studio | Dbeaver
• AI Modelling / Techniques:
• ▪ LLMs (LangChain)
• ▪ Computer Vision
• ▪ NLP (BERT)
• ▪ Transfer Learning (VGG16)
• ▪ Graph Analytics Community
• Detection, Egonet, Graph Embedding
• ▪ XGBoost, Random Forest, Baggers
• ▪ GANs, AutoEncoders, Isolation
• Forest
• ▪ SVM, SVR, NN, RBF, SOM, TDNN
• ▪ CNN, RNN, LSTM
• ▪ K-means/Birch Clustering/ GMM
• ▪ KNN, PCA, KPCA
• ▪ Logistic regression, GLM, Isotonic,
• Lasso and Ridge regression
• Others:
• ▪ Arc GIS | QGIS
• ▪ Scraping (Scrapy, Selenium)
• ▪ Visualization (Matplotlib, Seaborn,
• Plotly)
• ▪ Tableau | Power BI | Qlik
• ▪ Git |Jupyter | Jira/Ms Devops

Jan 2009 – Dec 2009 · 11 mos · Amsterdam Area, Netherlands

Jul 2003 – Dec 2008 · 5 yrs 5 mos · Delft Area, Netherlands

• Built 20 prediction models targetting asphalt damages for maintenance purposes. This was built for all asphalt types available in the Netherlands, ZOAB (Zeer open asfalt beton), DAB (Dichte Asfalt Beton) and SMA (Steenmastiek Asfalt).
• The project might drastically cut road maintenance costs, which are about 180 million euros annually. After building a GUI the models are used by Dutch Ministry of Transport and Water Management and road contractors.
• Produced research proposal that secured € 175,000 funding from Rijkswaterstaat (RWS/DWW) (2003-2004)
• Wrote 24 International Publications (2004-2009)
• Received 5 International Awards and Grants (2006-2009)
• Used and combined SVR, NN, RF and RST models to built regression and classification models. This study resulted in 20 intelligent models and eight different feature selection methods were applied including genetic serach.
• Did invited talk for Gent University, Faculty of Computer Science and Applied Mathematics, Gent, Belgium (2006)
• Passed 8 courses successfully and a number of workshops
• Presented 15 research papers in national and international conferences
• Was program committee member/reviewer of 6 journals and conferences