• Developed a Django-React web application for calculating deck material list, component dimensions, and rendering 3D designs based on user input.• Developed a web application for calculating dimensions of deck component parts and cost given user input deck dimensions (Django, Python, HTML, CSS, React.Js, PostgreSQL, AWS Lambda, APIGateway).• Develop a web application for rendering the deck design structure in 3D after obtaining the deck component dimensions given user input deck dimensions• Developed backend services for streaming video files, creating compressed zipped files, and pre-signed URLs using Amazon SQS, S3, Lambda, DynamoDB, and Python automation.• Offer two-course modules on Design Patterns, Data Structures, and Algorithm as well as Linux Operating Systems and Networking to college students at the Canada College of Technology and Business

-> Built financial model web services and APIs that allow access to company institutional and fundamental financial models and data in preference to spreadsheets.-> Maintained hundreds of thousands of financial models that the world's largest, most successful investment managers rely on to get the cleanest fundamental data and leverage their time.-> Developed API backend databases, services, and applications that enhance investment managers’ access to fundamental financial models by less than 80% compared to the conventional spreadsheet approach-> Participated in the conceptualization, designing, building, and upgrading of various products and services by the engineering team-> Developed software services using Django libraries that process over 4000 company model workbooks to extract useful financial data; store them in the database and render the data as an API service-> Automated the processes for cloning the production database to review app using the Amazon Web Services (SDK) as production data run on PostgreSQL installed on the AWS Relation Database Services-> Continuous Integrated and deployed source code using the Gitlab SDK, Gitlab pipeline jobs, Python, and bash shell scripting-> Scripted GitLab pipeline jobs that automate testing, data population to applications, staging, and deployment of application features on Heroku-> Mentored fresh graduate developers and intern/co-op on acceptable techniques in developing production-ready software-> Documented the team's method of building, maintaining, monitoring, securing and deploying software products and services-> Optimized API performance to ensure queried data (70 million records) are rendered in less a second to enhance API user satisfaction experience by 80%-> Gave presentations on completed features and tasks to a small and large technical audience within the company-> Collaborated with product and other team members to ensure tasks and projects are completed within the target time

Institutional quality financial models and fundamental data — at the click of a button — and the tools you need to integrate them seamlessly into your research workflow. We build and maintain fundamental models that the world's largest, most successful investment managers rely on to get the cleanest fundamental data and leverage their time. The company was acquired by Tegus Incorporation in August 2022.

1) Developed an Optical Character Recognition (OCR) RESTful microservices web application for converting uploaded Pdf to Docx files. The tech stack included Sanic (async flask for the main server), Django (Admin server), RabbitMQ (Event Bus messaging), MySQL, Docker. The deployed application preserves the pdf data object format in the Docx documents and reduced Docx file production cost by over 80% relative to alternative means.2) Developed a RESTful Question-Answering Machine Learning App using python packages, Elasticsearch SDKs, Google BERTs (LLM), Haystack, Hugging Face models, Docker that enables context information querying from pdf documents more efficiently, accurately, and faster by over 80%. 3) Built an API service that automates the extraction of email data objects and merged the extracts as pdfs documents by the SMTP libraries, Thunderbirds, python libraries, pdf plumber, and TDD, which reduces manual labor that takes 25 persons a year to less than 4 days. 4) Developed various software applications based on the variant Python web server frameworks Sanic(fast async)/Flask, Django, PostgreSQL, Firebase, Insomnia/Postman, Redis, Pytest, Github, that scales-up the legal business and reduce overhead running business cost by over 60%. 5) Automated a Contract Management System using Python web server frameworks (Sanic, Django), relational DB (PostgreSQL), and nonrelational DB (Firebase) that reduce the time taken to issue a contract by over 70%.6) Deployed applications on GCP/EC2/Heroku cloud environment given their benefits of security, scale-up, easier software version releases, and adequate infrastructure.

Backend digital insurance mobile/web advisor development Notable Contributions:1) Designed, developed, and maintained various insurance policy issuing processes in the app that enabled the business to have B2B white-labeled app. Backend services were also developed for obtaining optional insurance covers (e.g. Car Hire and Accidental Damages) for all listed assets in the mobile app using Sanic, Firebase, Redis, Lambda, SqS deployed on an EC2 instance. The user experience was enhanced using Locust.io libraries to test user concurrent app requests and server load capability. 2) Developed various REST APIs server-side endpoints using Sanic web framework for emailing client quotes. Policy schedules, wordings, and documents using Javascript, AWS serverless S3, lambda, and SeS. The APIs were documented using Sanic Open API swagger that enabled over five different B2B business deals within two months. The app functionalities enhanced user experience given the server speed response that increased by a factor of 5. 3) Developed an insurance quote recommendation engine (Cherry Picker) based on user’s preference on the amount of premium, essential cover, convenience cover or excess, 4) Automated some insurance risk factor code mapping, API mapping, and database migration as well as developed backend source code that integrates the company's REST API with various providers and third-party SOAP APIs.Utilized tools are Python 2+ & 3+, REST and SOAP APIs development,Swagger (Sanic Open API) for REST API's documentation,Postman, Insomia (endpoint testing), Sanic, Flask & Django Web app framework and Server, Server-side template: Jinja2, JavascriptDatabase: Firebase NOSQL,SQL,PostgreSQL Cloud VCS: Bitbucket, gitlab, github,Automated Test: Unittest, Mock test,, Pytest, TDDCloud Deployment: AWS (EC2, S3, SeS, SqS, lambda), Docker, microservices,Messaging: Slack, IntercomIn-memory database and cache: RedisOS:Linux Ubuntu

1) Automated web scraping process using Beautifulsoup, Selenium, Splinters to build data sources and pipelines for advanced descriptive analytics. Subsequently, developed and deployed Machine Learning (ML) model for predictive analytics of cryptocurrency portfolio management, risk analysis, and backtesting as required for strategic investment, 2) Automated the data cleaning, data feature reduction (supervised, unsupervised learning, reinforcement learning, optimization, and Bayesian inference), data feature ranking, using Extratree regressor - sci-kit-learn and Tensorflow libraries, as required for model training. 3) Developed and refactored a Financial Trading bot based on Deep Neural Network Learning (RNN, LSTM, Encoders, Embedders, GCP, Tensorflow, Keras) that yielded over 70% prediction accuracy and generated huge company revenue from gainful trades. 4) Developed live Sentiment Analysis Service by the stack of Python, Elasticsearch, Kibana, Grafana, REST APIs & dynamic scripted dashboards that increases the trading bot prediction accuracy by 71% when included in the model features (explanatory variables) 5) Conducted research studies on Initial Coin Offerings (ICOs), cryptocurrencies, their valuation, and factors influencing their valuation. Accordingly, I wrote sections of various published articles including an investment fund whitepaper and medium articlesUtilized tools for development in Python are: IDE:-> PyCharm, Eclipse, Texteditor & Jupyter Version Control System:-> Github, Gitlab Bitbucket Dataframes and Databases:-> Python Pandas & Elasticsearch (SDK), Logstash (EL), Google Bigquery, SQL and Mnesia (Erlang), Firebase (NoSql), Redis, Pyspark (distributed cluster data processing) Visualization:-> Kibana and Grafana Agile tools:-> Jira, Confluence, GitLab, Slack(SDK), Telegram, Intercom

1) Analyzed industry process and operational data (metallurgical mining furnace data) allowing the demonstration of competence in data preparation, imputation, visualization, feature selection, and reduction, 2) Explored the Principal Component Analysis (PCA), RFE, Extratree methods for the feature reduction while pandas packages and Scikit-learn libraries were utilized for model creation and validation. The predictive analytics method was utilized to determine the cause of unusual blowback phenomena events (positive pressure events) in the Polokwane furnace in South Africa.

1) Developed an Erlang application for querying the company's log directories and files of transaction activities using a pre-selected querying parameter. The tech stacks were Mochiweb (Erlang Web Framework) and Mnesia (database). I studies the book: "Learn You Some Erlang for Great Good" authored by Fred Hebert to master the Erlang syntax as invented by Ericson for phone switches. Erlang /OTP code execution is by pattern matching implemented by functional and concurrent programming language.

Developed a python software package that loads randomly ordered event sessions with allotted times and sort the event headings into sessions with the allotted times translated into AM and PM time format. The script also arranges the event sessions according to time progression with the lunch and break sessions included.(Company & Date: Jumo World - April 2017)Role: Python DeveloperDeveloped a python package that calculates the aggregate loans by the tuple of (Network, Product, Month) with the total currency amounts and counts written as “.csv” output file. The program was unit tested successfully and implemented as in a production environment(Institution & Date: African Institute for Mathematical Sciences, 2009)Role: Python Scientific Software Developer Developed a python package for digital deconvolution of Seismic Waves using Fourier Forward and Inverse Transform (Taught MSc Essay)Role: Python DeveloperDeveloped Python (2.7 and 3.4) packages that read randomly generated IP addresses from a file and sorts out the valid and invalid IP addresses; writes the valid and invalid IP addresses in descending order into separate output files. The package was unit-tested successfully and has been integrated into a web production code