Abdul-Baaqi Jempeji

AI in Capturing Properties of Textile Industry Using Deep Learning Models

Collaborated on designing a tool that permits simultaneous learning of the dictionary and encoding of extracted features based on the features extracted from different convolutional layers. This tool can be implemented using the “Learnable Encoding Module (LEM)” or a similar one.

Classifying Online Content to Provide Sustainability Advisory for Small Businesses

Collaborated on building a ML solution to classify and process the data provided by the client (raw text, extracted from research papers and other mediums) so that the relevance, format, authority of the source, etc. can be understood (Content Classification) and the content of the extracted sources can be understood and tagged for further processing and formatting, by either a human or Natural Language Generation (Content Processing).

Feasibility and ROI Analysis for Renewable Resources Infrastructure using Computer Vision

Collaborated on designing an AI solution to install Renewable Energy in Germany in order to speed-up the country’s course towards Net-Zero carbon emissions. Through this project, we were able to project the availability of renewable resources within the country including wind power, solar power, and hydro power. Do the feasibility analysis for the cost which will be incurred to set-up these prospects and finally the ROI of the same. This can then be used not only by the government but also by private entities to look for potential opportunities towards clean and green Germany. This will also help to accurately estimate the timeline toward reaching the goal instead of a vague 2045 Deadline.

AI for Road Safety

Collaborated on designing an AI solution to detect driver’s activities like driver talking on the phone, eating while driving the vehicle, detect drowsiness, distraction, yawn, eye closure and alert him when potential drowsiness situation is detected, from driving style like sudden accelerations or decelerations, sudden braking, sharp turns, set of events like start, stop, speed and turns, Maximum and minimum rpm of the engine, etc., The project results will be made open source. The aim is to help ride-sharing service providers, regulatory bodies, policymakers, and people while educating aspiring data scientists in solving real-world problems.

Detect Harmful Situations in Videos To Protect Children

Collaborated on building a computer vision model that is able to detect harmful and dangerous situations in video challenges that could go viral online. The AI solution classifies activities that could cause self-harm in a video, predicts if there is the possibility of harm to others in the video and classifies if the video encourages others to participate or not. If there is a high potential for harm and going viral, then the video can be classified as a dangerous challenge.

Improving Water Supply Sustainability in Nigeria Through Artificial Intelligence

Collaborated on designing an AI solution for Nigeria which uses GIS to help government and non-governmental organizations to be able to identify areas lacking clean water supply, provide easy access to water supply services, and be able to detect non-functional water supply services.This serves as a means to better the lives of Nigerian citizens.

Flower Images Classifier

Implemented an image classification application that trained a a deep learning model on a data set of images, then use the trained model to classify images.

Automatic Number Plate Detection

Implemented an Automatic number plate application. The model identifies the plate from the image, extract the characters and cross-checks it with the database to retrieve information as regards the owner of the plate number.

Finding Paths to Safety Following Natural Disasters with Satellite Imaging and AI

Collaborated on building a model that collects data about the local roads from satellite images, classify them and indicate the safest route to be taken from point A to point B, displays an interactive dashboards to display road characteristics and safest paths and then arrange demonstration and publicise to local audience.

Dog Images Classifier

Used a created image classifier to identify dog breeds while correctly identifying which pet images are of dogs (even if breed is misclassified) and which pet images aren't of dogs and correctly classify the breed of dog, for the images that are of dogs.

Using Computer Vision to Estimate Biomass of Fish for Sustainable Food Production

Sustainable Development Goal 14 (Goal 14 or SDG 14) is about “Life below water” that aims to conserve and use sustainably the oceans, seas, and marine resources for sustainable development. In this project, we built an ML algorithm that estimates the volume and weight of a fish from a video or an image, as well as a procedure to calibrate the system.

User Response prediction

My submission for Data Science Nigeria 2021 AI Bootcamp Qualification Hackathon hosted on Zindi. The goal of the hackathon is to build a model capable of predicting users’ responses to marketing campaigns based on the features in the provided dataset by the sapa.com data engineer.

Machine Learning Engineer

• Leveraged computer vision to estimate biomass of fish for sustainable food in line with SDG 14 thus decoupling the production of animal protein from fishing and traditional agriculture and releasing the pressure on oceans and other ecosystems that are already strained.

Junior Machine Learning Engineer

• Leveraged computer vision to estimate biomass of fish for sustainable food in line with SDG 14 thus decoupling the production of animal protein from fishing and traditional agriculture and releasing the pressure on oceans and other ecosystems that are already strained.
• Used computer vision to improve water sustainability which enables the government and non-governmental organisations to identify areas lacking clean water supply, provide easy access to water supply services and to detect non-functional water supply services in Nigeria.
• Designed an automatic number plate detection system which is deployed by users namely schools, estate, government agencies and companies to track, match and identify vehicles, hence promoting security of lives and properties
• Developed a technological system using satellite imaging and Artificial Intelligence to provide paths to safety following a natural disaster which is used by families to develop a disaster response plan.
• Built an AI technology that detects and filters harmful situations in online video contents thus protecting children.

Faculty

I teach different courses, majorly, Islamic FInance, Economics and Accounting.

Management Trainee

• Extracted data from internal and external databases; analysed exploratory data; cleaned, transformed, and aggregated datasets which eliminated errors, provided high quality information and improved productivity.
• Prepared data analysis and visualisation using excel dashboard, tableau and Power-BI to prepare company’s data on fertilisers, finance, human resources among others which enabled better understanding of data and aided decision making.
• Built tools to automate data collection which reduced errors, saved time, eased work process, enhanced productivity and improved business efficiency.
• Developed, maintained, and improved holistic datasets, reports and dashboards which help monitor and improve the company’s performance.
• Engaged in cross-divisional collaboration to analyse data and present results which were used to improve financial performance.

Data Analyst Intern

• Cleaned and analysed records for 5,000,000+ commercial airline flights in 2015, compiled for the U.S. DOT Air Travel Consumer Report.
• Cleaned and analysed historical Stock Price of 10 Popular Companies (Apple, Amazon, Netflix, Microsoft, Google, Facebook, Tesla, Walmart, Uber and Zoom).
• Cleaned and analysed New York City Taxi Trips containining over 28 million Green Taxi trips in New York City from 2017 to 2020.
• Cleaned and analysed restaurant ratings in Mexico by real consumers from 2012, including additional information about each restaurant and their cuisines, and each consumer and their preferences.
• Cleaned and analysed Shark attacks reported over the past 100 years, including location, activity, victim info (name, gender, age), shark species, etc.

Research Assistant

• Conducted an extensive research on accessing the inflation hedging of gold and palladium in OECD countries which increased my knowledge and research skills in this field.
• Supported my supervisor in preparing detailed research on the topic “The inflation hedging properties of gold, stocks and real estate: A comparative analysis”.
• Prepared a research work on new evidence for the inflation hedging potential of US stock returns which improved my commercial awareness and enabled me to develop a global mindset.

Audit Assistant

• Prepared financial reports which ensured compliance with tax regulations; monitored the accuracy of accounting systems and procedures and ensured compliance of the firm with laws and international best practices.
• Reviewed, assessed and made valuable recommendations which ensured changes were constantly made to business units accounting system and control.
• Prepared and maintained audit budgets, planned and coordinated the company’s financial audits which promoted financial transparency.
• Collaborated with other members of audit and accounting teams in performing their daily tasks all geared towards clients’ satisfaction and increasing profitability of the firm.

Research Assistant

Carried out Research.

Udacity

International Centre for Education in Islamic Finance (INCEIF)

University of Ibadan