I am Ashaba Joshua Jasper, a data scientist, AI/ML engineer and full-stack developer in Kampala, Uganda. I design and build practical systems end to end: data pipelines and models, REST APIs and full-stack web applications.
My work sits where data science, applied AI and software engineering meet. I care most about tools that solve real operational problems for businesses, co-operatives and non-profits in East Africa: the hotel that needs one set of books, the firm that needs to see its deadlines, the public that needs its data in a form it can use.
In short
Data scientist and AI/ML engineer with a First Class computer science degree and a record of shipping production data products end to end: data pipelines and LLM systems, observability dashboards and full-stack applications. Most recently delivered AI/ML, data-science and analytics systems for a leading US radiology network (Envision Radiology, through Reveloop Tech Systems), spanning multi-LLM clinical analysis, document-intake automation, DICOM computer vision on Kubernetes and Grafana telemetry across the RIS and PACS ecosystem. Combines metric-driven analysis with the engineering discipline to put models into production and keep them observable.
What I do now
I am co-founder and COO of Persmon Technologies (opens in a new tab), a software company in Kampala. My work with Persmon dates to December 2022. Our portfolio includes 47 projects, from a hotel management system and an operations platform to an online bookshop and a civic data platform. The full list is here.
I am also building Learnnovate (opens in a new tab), a technology education non-profit, where I have been Founder and Program Director since December 2022. Its programmes have reached 200+ learners across multiple schools.
Before Persmon, and alongside it
From February 2025 to June 2026 I worked remotely as a data scientist and AI/ML engineer with Reveloop Tech Systems for Envision Radiology, a US radiology network. There I built RadCareLoop, a multi-LLM follow-up analyser (Gemini, GPT and Claude with a Judge arbitration layer) validated at 97.9% accuracy against a 10,000-report gold-standard dataset, a document-intake pipeline for fax orders, DICOM computer vision on Azure Kubernetes Service and Grafana telemetry for the RIS and PACS ecosystem.
Earlier roles took me through data engineering and BI at Uganda Bookshop, e-commerce and analytics at Excellent Shop, Flutter apps at Centenary Publishing, predictive maintenance for fleets at Blue Pearls and Uganda Transporters, and an internship in operations and monitoring at MTN Uganda. The full CV has the dates and results.
What I am learning
Right now I am going deeper into MLOps, data engineering and production-grade applied AI: the part of the work where a model stops being a notebook and starts being a system other people depend on.
How I work
For data and models I work in Python with PyTorch, TensorFlow and scikit-learn, on PostgreSQL, MySQL and SQLite. Applications are mostly TypeScript and JavaScript with React, Next.js and Tailwind CSS, with services in FastAPI, Django, Spring with Java, and Node.js. I ship with Docker, GitHub Actions, Linux, Bash and Git, onto Azure, Firebase, Supabase, Vercel and Netlify. For mobile I reach for Flutter and Dart, and for reporting, Power BI and Tableau.
On GitHub
Some smaller projects I have built in the open:
- PaveStruct (opens in a new tab), a pavement design tool for civil engineers that follows SATCC standards, with a six-step workflow and live structural-number calculations.
- SilverVibe SACCO (opens in a new tab), a demo management platform for a Ugandan savings and credit co-operative, with an admin dashboard, a member portal and role-based views.
- Bookstore Inventory API (opens in a new tab), a containerised REST service in Java and Spring Boot for books, authors, categories and orders, built around a layered architecture.
- SMART Debt Tracker (opens in a new tab), a privacy-first personal finance app that compares snowball and avalanche payoff strategies, with all data kept in the browser.
- Handwriting Recognition (opens in a new tab), a neural network that classifies handwritten letters, with OpenCV preprocessing and notebook-based evaluation.
- Medical Diagnosis with Bayesian Networks (opens in a new tab), a probabilistic model that reasons from observed symptoms to likely diagnoses.
Education
I studied at Uganda Christian University from January 2022 to July 2024, where I earned a BSc Computer Science, First Class Honours with a GPA of 4.62/5.0. My capstone, Help Anonymous, was an AI-powered mental-health app with BERT and LSTM sentiment analysis that reached 270+ active users.
I was Lead of the Google Developer Student Club from 2022 to 2023, taking 100+ students through workshops on machine learning, Python, TensorFlow and cloud, and General Secretary of the Data Science Society. In the same period I was ranked among Africa’s Top 100 in the Google Developer Community for Machine Learning and Cloud.
Recognition and certifications
- Gold Award, Queen's Commonwealth Essay Competition (2021)
- First Class Honours, BSc Computer Science (GPA 4.62/5.0), Uganda Christian University (2024)
- Top 100 in Africa, Google Developer Community for Machine Learning and Cloud, Google Developer Community (2022 to 2023)
- I4G Cybersecurity Scholarship, Cisco Networking Academy (2023)
Certifications include:
- IBM Data Science Professional Certificate, Coursera (2026)
- Machine Learning Specialization (Andrew Ng), DeepLearning.AI, Coursera (2026)
- Junior Cybersecurity Analyst and Cyber Threat Management, Cisco Networking Academy (2023)
- Introduction to Machine Learning, Kaggle (2023 to 2025)
- Google Cloud Essentials, Google Cloud Skills Boost (2023)
- Digital Marketing Professional Certificate, CertiProf (2023)
Get in touch
I am open to collaboration on data, AI and software projects. The fastest way to reach me is email or LinkedIn (opens in a new tab), or you can use the contact form.