Saad is available for hire
Hire SaadSaad Kayyali
Verified Expert in Engineering
Data Scientist and Data Engineer Developer
Location
Amman, Amman Governorate, Jordan
Toptal Member Since
October 5, 2022
Saad is a data scientist and developer, passionate about tackling challenges that require learning and thinking outside the box. He has over a decade of experience and an academic background in electrical engineering and mathematics. Saad has worked with large corporations and small startups and is able to lead a team of engineers, as he did at Samsung, or start a new data department like at Jordan's Capital Bank.
Portfolio
Capital Bank of Jordan
Informatica, TIBCO, Netezza, SQL, Python 3, TensorFlow, Pandas...
Self-employed
Django, Python 3, Ruby on Rails 4, PostgreSQL, Git, Front-end, Back-end...
Daman Health Insurance
SAS, Python 3, R, SQL, SAS Machine Learning, MATLAB
Experience
Availability
Part-time
Preferred Environment
Jupyter Notebook, Docker, Linux, Windows 11
The most amazing...
...challenge I've overcome was running an ML model that crashed supercomputers in Munich Re. I optimized and rewrote the code until it ran locally on my laptop.
Work Experience
Data Scientist
2020 - 2022
Capital Bank of Jordan
- Established the data management department from scratch. Worked with multiple teams to set up and design the infrastructure, built the ETLs with consultants, and did analysis and predictions after the new DWH was established.
- Built ETL pipelines on TIBCO Spotfire and Informatica. The data was extracted from multiple databases and/or static files. The pipelines were written in SQL or Python Pandas. Created calculated columns and built the final dashboards.
- Developed an ML model to match customer names with accounts in other banks. The aim was to analyze customer behavior, see if they send money to accounts in other banks, and tempt them to keep the money. Worked with Arabic and English NLP.
- Built a recommendation engine using TensorFlow Recommenders and ranking per customer clusters based on a collaborative filtering two-tower model. It recommends stores for customers based on their shopping behavior and other similar customers.
- Created our own Python library and hosted it internally in our own Anaconda Server to be used across the team. It has the most common functions we use and is optimized for fast data extraction.
Technologies: Informatica, TIBCO, Netezza, SQL, Python 3, TensorFlow, Pandas, Oracle Data Integrator 11g
Technology Expert
2017 - 2020
Self-employed
- Integrated the front-end and design team in developing Pandia Health, a startup in California that links women with doctors to deliver birth control solutions.
- Built a customer-facing tool for StratticaLabs, a startup in California that offers loan computer-assisted decisions for financial institutes. I was part of the data analytics team.
- Co-founded Living Mosaic, an architecture social startup that repurposes unused spaces in the city into beautiful spaces for the community.
Technologies: Django, Python 3, Ruby on Rails 4, PostgreSQL, Git, Front-end, Back-end, Data Analysis
Data Scientist | R&D
2015 - 2017
Daman Health Insurance
- Built an ML model to check the validity of medical claims that links diagnoses with diseases. The model was able to process claims with 87% accuracy. We used various NLP research papers like Med2Vec.
- Analyzed millions of claims to find patterns and anomalies. The framework allowed us to predict prescription abuse by doctors. It detected Vitamin B-12 abuse.
- Identified patients with a high risk of diabetes. We contacted those patients and coached them to change their lifestyle before getting the disease.
Technologies: SAS, Python 3, R, SQL, SAS Machine Learning, MATLAB
Co-founder
2010 - 2017
Google Developer Group - Dubai Chapter
- Created one of the biggest tech events in Dubai, such as Droidcon-Codemotion and Devfest. I managed the event from idea to execution. We got speakers from around the world and delivered conferences for hundreds of attendees.
- Delivered workshops and talks in many countries and universities, including Middlesex University and the American University of Sharjah. I taught design sprints, Android development, and machine learning.
- Got trained by Google on their technologies like TensorFlow and how to be a trainer. Furthermore was trained in design sprints, design thinking, and ideation.
Technologies: TensorFlow, Design Sprints, Android SDK, Google Glass, Event Management, Design
App Content Manager
2008 - 2012
Samsung
- Acted as the team lead for test engineers across MENA. We tested the hardware, software, and apps for Samsung phones before being released on the market.
- Trained developers and retailers on the phones' new functionalities and how to sell them or develop apps for them using Samsung APIs as the product trainer and technical evangelist for the MENA region.
- Got voted as the most creative engineer, and one of my app designs became native to the Android operating system.
Technologies: Bada, Java, Microsoft Office
Experience
Recommendation Engine
The Capital Bank of Jordan wanted to incentivize their clients spending in general and specifically through credit cards. After multiple brainstorming sessions, we decided to analyze customers' spending habits and recommend stores and places to go shopping. To do that, I built a model using TensorFlow that suggests stores for customers depending on their shopping history. The final product sends a template email to the selected customers with personalized suggestions.
New Data Warehouse
The Capital bank of Jordan needed a new data warehouse. The main challenge was dealing with all data sources used by the bank, such as T24, the core systems for financial transactions, ATM database, mobile app, website, Iraq branches database, and IT logs. We created ETL tools that extracted, transformed, and loaded the needed data into our new Netezza database.
Diabetes Prediction Model
While working with Daman Heal Insurance in UAE, our team built a model that can predict pre-diabetic patients and do an intervention to change their lifestyle and protect them from becoming diabetic. I worked on the prediction model and data processing.
Skills
Languages
Python 3, SQL, SAS, R, Java
Libraries/APIs
Pandas, Scikit-learn, TensorFlow
Tools
Plotly, IBM Watson, MATLAB, Git, Informatica ETL
Paradigms
Data Science, ETL
Platforms
Jupyter Notebook, Docker, Linux, Oracle Data Integrator 11g, Google Glass, Google Cloud Platform (GCP)
Other
Machine Learning, Informatica, Data Engineering, TIBCO Spotfire, Windows 11, TIBCO, Wireless Communication, Mathematics, SAS Machine Learning, Front-end, Back-end, Data Analysis, Bada, Microsoft Office, Design Sprints, Event Management, Design
Frameworks
Selenium, Django, Ruby on Rails 4, Android SDK
Storage
Netezza, PostgreSQL
Education
2003 - 2007
Minor in Mathematics
American University of Sharjah - Sharjah, United Arab Emirates
2003 - 2007
Bachelor's Degree in Electrical Engineering
American University of Sharjah - Sharjah, United Arab Emirates
Certifications
DECEMBER 2019 - PRESENT
Google Cloud Platform Fundamentals: Core Infrastructure
Google Cloud | via Coursera
DECEMBER 2019 - PRESENT
Essential Google Cloud Infrastructure: Foundation
Google Cloud | via Coursera
NOVEMBER 2019 - PRESENT
Getting Started with Google Kubernetes Engine
Google Cloud | via Coursera
NOVEMBER 2019 - PRESENT
Essential Google Cloud Infrastructure: Core Services
Google Cloud | via Coursera
NOVEMBER 2019 - PRESENT
Elastic Cloud Infrastructure: Containers and Services
Google Cloud | via Coursera
NOVEMBER 2019 - PRESENT
Elastic Google Cloud Infrastructure: Scaling and Automation
Google Cloud | via Coursera
NOVEMBER 2019 - PRESENT
Reliable Google Cloud Infrastructure: Design and Process
Google Cloud | via Coursera
Collaboration That Works
How to Work with Toptal
Toptal matches you directly with global industry experts from our network in hours—not weeks or months.
1
Share your needs
Discuss your requirements and refine your scope in a call with a Toptal domain expert.
2
Choose your talent
Get a short list of expertly matched talent within 24 hours to review, interview, and choose from.
3
Start your risk-free talent trial
Work with your chosen talent on a trial basis for up to two weeks. Pay only if you decide to hire them.
Top talent is in high demand.
Start hiring