Ayobamidele.
PhD student in Computer & Software Engineering at Polytechnique Montréal. Building scalable software systems and researching formal verification of safe, trustworthy AI.
About.
My journey from Electrical Engineering to Fullstack Development and Research.
Currently
Building: Local-first AI systems with verifiable safety guarantees
Researching: Formal verification of learning-enabled autonomous systems
Reading: Neuro-symbolic AI & hybrid verification methods
3
Publications
6
Projects
5+
Research Domains
4
Countries
I build backend-first fullstack applications with Python and modern frontends — while pursuing research on formal verification of learning-enabled autonomous systems, safe AI, and LLM-assisted software engineering.
Journey
PhD in Computer & Software Engineering
Formal Verification & Software Engineering for Safe Autonomous AI Systems. Supervised by Prof. Lina Marsso.
Research Intern
LLM-driven pipeline using the MIT Deliberatorium for clustering, evaluating, and summarizing deliberative arguments using MECE-aligned logic.
MSc in Collective Intelligence — Valedictorian
School of Collective Intelligence, Rabat, Morocco. Graduated as Valedictorian, Class of 2025.
Research Member
Empirical research on crowd-based bias in sentiment analysis and medical outcome prediction. ML pipelines for breast cancer recurrence prediction.
Software Engineer
Developed scalable educational software platforms; applied software engineering practices for dependable and maintainable systems.
B.Eng. Electrical & Electronics Engineering
CGPA: 4.10/5.00. Thesis: Optimal Placement of UPFC and STATCOM for Voltage Stability Using ANN Modeling.
Projects.
Backend-first, local AI systems built for real-world impact.
Socratic Depth Engine
A tutor that never gives answers — only better questions
AI tutors explain things. This one refuses to. It builds a live causal knowledge graph of what a learner actually understands, then finds the single minimum question that unlocks the next concept.
Neighbourhood Pulse
The stress map your city doesn't know it needs
City health departments allocate mental health resources based on 5-year-old census data. This platform ingests real-time anonymous signals — 311 calls, subreddit patterns, food bank demand — to produce a living neighbourhood stress index.
Deliberation OS
Turning 10,000 community voices into one actionable brief
Public consultations collect thousands of comments, then a consultant summarises them with their own biases. This platform clusters raw input into a live argument map, surfaces hidden consensus, identifies genuine disagreement, and auto-generates a policy brief.
Cognitive Mirror
Your entire intellectual life, made legible to you
Thousands of notes, journals, and documents — buried in them are contradictions you've never noticed, ideas you forgot, and knowledge gaps you keep circling. This tool ingests everything locally and builds a personal knowledge graph that surfaces these patterns.
Clinical Narrative Intelligence
Turning doctor notes into patient journeys a human can see
A patient's history is thousands of unstructured notes in jargon and abbreviations. No one reads them all. This system extracts a structured timeline, flags drug interactions and care gaps, and renders a visual narrative navigable in 60 seconds.
Crisis-to-Care Bridge
The gap between 'I need help' and 'here is help' should be zero
Someone in crisis searches for help and gets closed hotlines and outdated numbers. This platform runs real-time NLP triage — classifying urgency, need type, and barriers — then matches to a live, verified, hyper-local resource graph maintained by social workers.
Research.
My research interests, publications, and contributions to the academic community.
Research Interests
Formal Verification of AI
Formal verification of learning-enabled and autonomous systems — ensuring AI behaves correctly by mathematical proof, not just testing.
Safe & Trustworthy AI
Robustness and interpretability of neural networks. Building AI systems that are reliable, explainable, and safe for deployment in critical domains.
Neuro-Symbolic AI
Hybrid AI systems combining neural learning with symbolic reasoning for verifiable decision-making pipelines.
LLM-Assisted Engineering
Using large language models to assist software engineering tasks — specification inference, code verification, and automated reasoning.
Collective Intelligence
Computational methods for aggregating and structuring human deliberation at scale — from argument mining to consensus detection.
AI for Healthcare
Machine learning pipelines for clinical prediction, medical NLP, and patient outcome modeling with emphasis on responsible AI practices.
Publications(3)
Effects of Clinically Designed Improvisory Music on Learning Outcomes and Stress
P. Adubi, T. Sanda, R. Oyewole, R. Akodu, C. Oyewale, S. Omotunde
Proceedings of ICERI 2024, IATED · 2024
A Comparative Study of Machine Learning Models and Imputation Techniques for Predicting Breast Cancer Recurrence
P. Adubi, S. O. Samuel
Manuscript under preparation · 2024
Optimal Placement of UPFC and STATCOM for Voltage Stability in Power Networks Using ANN Modeling
P. Adubi, O. I. Adebisi
Manuscript in progress · 2020
Skills.
Technologies and tools I work with daily.
Programming
Python
PrimaryJavaScript
PrimaryR
ProficientMATLAB
ProficientTypeScript
ProficientGit
PrimaryMachine Learning & AI
Neural Networks
ProficientDeep Learning
ProficientXGBoost
ProficientRobustness Analysis
FamiliarModel Interpretability
FamiliarLLMs & NLP
HuggingFace
ProficientGPT APIs
ProficientRAG Pipelines
ProficientSemantic Clustering
ProficientArgument Mining
FamiliarSoftware Engineering
FastAPI
PrimaryDjango / DRF
PrimaryReact / Next.js
ProficientDocker
ProficientPostgreSQL
PrimaryML Pipelines
ProficientFormal & Experimental Methods
Experimental Design
ProficientRCTs
ProficientStatistical Testing
PrimaryOptimization
ProficientFrameworks & Tools
PyTorch
ProficientTensorFlow
FamiliarScikit-learn
PrimaryOllama
ProficientNeo4j
FamiliarChromaDB
ProficientWorkshops & Talks.
Presentations, workshops, and talks I've given.
Poster Presentation
World Tech Summit on Big Data, Data Science & Machine Learning
2024
Presented research on machine learning approaches for predictive modeling in healthcare and education domains.
Poster Presentation
ICERI Conference 2024
2024
Presented findings on the effects of clinically designed improvisory music on learning outcomes and stress.
Poster Presentation
European Society for the Cognitive Sciences of Music
2024
Presented research at the intersection of music cognition, clinical music design, and educational outcomes.
Advanced Workshops on Research Data Science
International Centre for Theoretical Physics (ICTP)
2024
Intensive workshop on advanced data science methods for research applications, hosted by ICTP.
UM6P Science Week Participant
UM6P Science Week
2024
Participated in the university-wide science showcase featuring cross-disciplinary research presentations.
MENA Machine Learning Winter School
MENA ML Winter School
2025
Intensive machine learning training program bringing together researchers from the MENA region.
Blog.
Thoughts on engineering, research, and everything in between.
Contact.
Have a project in mind or want to collaborate? Let's talk.