AI Engineering Intern
Cytobiz Innovation Lab
Contributing to AI-enabled and data-driven product work. Selected product platforms include BetaRide, WeNaijaHealth, and AccessCare. Project links and product-level descriptions are provided below.
AI engineering · Applied machine learning · Data systems
Building useful
systems for real conditions.
I’m Quadri Adebisi, an AI Engineering Intern and B.Tech Statistics student at FUTMinna. I work across applied AI, backend engineering, data science, and product development — with a focus on systems that remain useful under real operational constraints in health, agriculture, education, mobility, and civic services.
AI Engineering Intern · Cytobiz Innovation LabPopularly Known As · A4 AnalyticsCTO · AmanKasuFUTMinna · Statistics
Applied AI · Engineering · Research
4.30/5.00FUTMinna GPA
70+NEXUS skills mapped
8AmanKasu target states
5WeNaijaHealth languages
Profile
Statistics, engineering
and applied problem solving.
My work is guided by a simple principle: a model or platform matters when it helps someone make a better decision or complete a task in their actual working environment.
I am completing a B.Tech in Statistics at the Federal University of Technology Minna and a Data Science programme at WorldQuant University. My foundation in statistics shapes how I approach evidence, modelling, evaluation, and uncertainty; my engineering work focuses on turning those methods into usable software.
My projects span low-connectivity disease surveillance, agricultural market intelligence, student career navigation, pharmacy operations, community issue reporting, institutional research management, health-insurance navigation, and local mobility. I have worked on systems that are live, in pilot, or at prototype stage; I distinguish those stages rather than presenting every build as a scaled deployment.
My current roles include AI Engineering Intern at Cytobiz Innovation Lab, Lead Advocate for Health Insurance at WeNaijaHealth, CTO at AmanKasu (formerly TradeFlow NG), and PKA at A4 Analytics. I also contribute to student learning and entrepreneurship communities through DSN and Hult Prize at FUTMinna.
Evidence before claims
Use metrics that have a clear source, context, and limitation. Separate field validation from full deployment.
Build for operational conditions
Account for unreliable connectivity, data quality, device constraints, privacy, and the people who will use the system.
Human-centred automation
Use AI to support analysis and workflows while preserving review and accountability where decisions affect people.
Experience & leadership
Roles, teams and
responsibilities.
Selected roles across engineering, health advocacy, applied data work, research, and student innovation.
2026–Present
Lead Advocate for Health Insurance
WeNaijaHealth
Support awareness and navigation for informal workers exploring health-insurance schemes, eligibility, costs, and official registration pathways. The platform supports five languages and is an independent access service, not an insurer or government agency.
Chief Technology Officer (CTO)
AmanKasu (formerly TradeFlow NG)
Provide technical direction for an agricultural trade-intelligence platform connecting field price collection, commodity forecasting, and route optimisation. Current public reporting describes 19 field agents across eight target states and field validation in three.
PKA
A4 Analytics
Build and deliver AI, data, and software solutions for operational problems, including pharmacy management, civic reporting, and institutional research intelligence.
Data Science & Statistics Resource Person (SIWES)
National Centre for Agricultural Mechanization (NCAM)
Developed AgriROS for research and prototype tracking, and supported statistical analysis of mechanisation trials, crop/seed physical properties, and adoption research.
Tech & Learning Support (Volunteer)
Data Scientists Network (DSN), FUTMinna
Designed a three-month Python-to-ML learning pathway for 52 students across eight departments using intake-survey data; automated parts of onboarding and community operations.
Campus Director
Hult Prize at FUTMinna
Lead the campus On-Campus Program, coordinating student engagement, startup-team development, mentors, and institutional partnerships.
Director of Research
Association of Statistics Students of Nigeria (ASSON), FUTMinna
Led 10+ research initiatives and coordinated data collection, statistical analysis, and formal presentation of findings. Received Academia of the Year and Executive of the Year recognition for 2024/2025.
Selected work
Projects with a clear
problem and purpose.
Each case study outlines the problem, how the system responds, what is known about its current stage, and my role. Links open the available product demonstrations.
Public health systems
Lafiya
Offline-resilient disease surveillance
Problem
Field surveillance and contact-tracing workflows can lose continuity when devices operate in places with intermittent connectivity. Data must be captured reliably and protected while it moves between devices and central systems.
Approach
I built the backend and data-engineering layer: schema-validated ingestion, privacy-preserving device identifiers, deterministic record IDs, deduplication, encrypted store-and-forward relay, and local SQLite persistence. Relay devices transport sealed bundles without reading the underlying health payloads. The resulting records feed downstream intelligence for exposure risk, vitals anomalies, contact networks, outbreak zones, and device health.
Evidence & stage
A substantial low-connectivity public-health engineering project developed with a three-person team. The initial implementation was built during HealthTrace Hackathon 2026, where the team placed first in System Design & Innovation. The recognition is one milestone; the work is presented here for its system design and public-health purpose.
Role: Backend and data engineeringPython · SQLite/PostgreSQL · REST APIs · encrypted relay · privacy-aware ingestion
Early field validation
Problem
Agricultural traders often make buy/sell and route decisions with fragmented price information. Surplus in one market may not reach a nearby market where demand and prices are stronger.
Approach
The platform combines weekly market-price collection by field agents, Prophet-based commodity price forecasts, and linear-programming optimisation. Recommendations consider buying and selling locations alongside transport cost, vehicle capacity, supply constraints, and risk. It is designed to support intra-state and inter-state market decisions as the data network grows.
Evidence & stage
The public project site reports 19 field agents across eight target states, physical field validation in three states, 12 trader interviews, and a 70-respondent demand survey spanning 16 states. These are validation and network-building indicators, not proof of national-scale impact.
Role: Chief Technology OfficerPython · Prophet · PuLP · PostgreSQL · Streamlit
Live platform · initial testing
Problem
Tertiary students can reach graduation without a clear view of how their current skills map to industry-facing career pathways, what gaps remain, or which opportunities are relevant to them.
Approach
NEXUS maps more than 70 skills across 11 technology career pathways. Students build a profile, compare current skills against a selected direction, explore matching opportunities, and receive practical next steps and learning roadmaps. The intent is to connect self-assessment with opportunity discovery rather than function as a passive job board.
Evidence & stage
A working platform built for FUTMinna students. Initial user testing has explored whether students can identify skill gaps and discover relevant opportunities in one sitting. Broader validation of recommendation quality and student outcomes remains a next step.
Role: Concept and product directionNext.js · TypeScript · PostgreSQL · Supabase · Python
Live in store
Problem
Pharmacy operations depend on accurate stock, pricing, sales, and customer accounts. Cloud-only systems become unreliable when a store loses internet access, while owners still need remote visibility.
Approach
A local-first operational system runs on local PostgreSQL and synchronises to the cloud in the background. It includes four-tier pricing, FEFO batch tracking, inventory and sales workflows, a customer debt ledger, role-based access, Google Forms stock-entry integration, and a remote CEO dashboard.
Evidence & stage
The system is live in-store. The design keeps core shop operations available offline while providing remote oversight through cloud synchronisation. It was delivered with setup documentation, Windows deployment scripts, and versioned database migrations.
Role: System delivery through A4 AnalyticsFastAPI · React · PostgreSQL · Supabase · Railway · Vercel
Deployed · pilot approved
Problem
Research publications, prototype records, and project milestones can be scattered across files, officers, and departments, making institutional monitoring and cross-team discovery difficult.
Approach
AgriROS provides role-segregated research tracking, structured document ingestion, LLM-assisted extraction and semantic search, cloud document storage, and a pending-approval workflow so human reviewers validate records before they become part of the institutional knowledge base.
Evidence & stage
Developed during my SIWES placement at the National Centre for Agricultural Mechanization. The source CV records deployment for NCAM and management approval for a five-week pilot involving the FPM and ESS departments.
Role: Platform developmentPython · FastAPI · Next.js · PostgreSQL · Supabase · Groq LLM
Working public-facing platform
Problem
Residents may not know which public authority should receive an infrastructure complaint, and often lack visibility into whether a report has been assigned or resolved.
Approach
CivicBot is a WhatsApp-oriented reporting platform for text and image submissions. The product combines AI-assisted issue classification with a live issue map, an administration dashboard for assigning reports and updating statuses, and analytics for tracking patterns and response progress.
Evidence & stage
A public-facing working platform is available. No deployment-scale, response-time, or issue-resolution figures are claimed here; those should be added only when supported by operational records.
Role: Product/system development through A4 AnalyticsPython · NLP · Computer Vision · Flask · PostgreSQL
Pilot corridor operations
Problem
Local commuters face fare uncertainty and fragmented access to commercial tricycles, motorcycles, and cars. Rigid surge pricing may not reflect local transport norms.
Approach
BetaRide supports benchmark-informed fare proposals, location-based dispatch to nearby operators, a single structured counter-offer, driver and vehicle checks, and direct passenger-to-driver settlement. The product is designed around locally defined corridors and transparent negotiation.
Evidence & stage
The public product site describes an active pilot corridor from Offa Garage Terminal to five strategic hubs in Ilorin and a 50-operator pilot fleet. These are pilot-stage indicators rather than evidence of citywide coverage.
Role: AI Engineering InternProduct contribution at Cytobiz Innovation Lab
Digital health · inclusion
WeNaijaHealth ↗
Multilingual health-insurance navigation for informal workers
Problem
Informal workers may struggle to understand health-insurance eligibility, costs, available schemes, and official registration routes. Language and process complexity can make access harder.
Approach
The platform provides a free, no-login, five-language navigation experience. Users share basic information about their work and location, explore relevant programme options such as KwaraCare and NHIA pathways, understand eligibility and cost considerations, and follow official registration steps. Team alerts and an operations triage console support navigation to verified clinic-enrolment pathways within a 24-hour response service window. It is an independent navigation platform, not an insurer or government agency.
Evidence & stage
The public website explains pathways to scheme information and support, including that coverage activation remains with the relevant scheme authority. My role is Lead Advocate for Health Insurance, supporting awareness and navigation. The 24-hour window is a service-response commitment, not a guarantee of insurance activation.
Role: Lead Advocate for Health InsuranceHealth insurance advocacy · multilingual access · digital health navigation
Live product link
Problem
Access to healthcare services can be difficult to navigate, particularly when people need to identify suitable pathways for care.
Approach
Selected product work at Cytobiz Innovation Lab. The public site is linked for review; this portfolio intentionally avoids describing specific workflows or outcomes beyond the product’s healthcare-access focus until the exact scope of my contribution is confirmed.
Evidence & stage
Project link is live. Product-specific impact figures are not included.
Role: AI Engineering InternProduct work at Cytobiz Innovation Lab
GovTech · responsible AI
EPIA
Ethical Procurement Intelligence for Africa
Problem
Procurement teams need better ways to review tender documents, identify compliance gaps, and spot patterns that may warrant human investigation.
Approach
The proposal combines offline-first procurement workflows, NLP-assisted tender review, graph-based anomaly detection concepts, alternative-data supplier assessment, and an audit-oriented dashboard. Model flags are intended as decision support, not findings of wrongdoing.
Evidence & stage
Shortlisted 69 of 240 entries in the ESG Student Innovation Challenge 2026. This recognition reflects the proposal’s competitiveness, not a deployed procurement outcome.
Role: Proposal and technical designPython · FastAPI · PostgreSQL · SQLite · NLP · graph analytics
Technical toolkit
Methods and tools
used across the work.
This is a grouped view of technologies and methods used in projects, rather than subjective proficiency percentages.
Programming & data
Python, SQL, R, JavaScript, TypeScript, Pandas, NumPy, Jupyter, Git, data cleaning, exploratory analysis, statistical inference.
Machine learning & AI
Supervised learning, regression, classification, NLP, computer vision, transfer learning, time series, LLM applications, retrieval-augmented generation, model evaluation.
Backend & databases
FastAPI, Flask, REST APIs, PostgreSQL, SQLite, Supabase, MongoDB, ETL, data validation, authentication, role-based access, row-level security.
Analytics & optimisation
Experimental design, hypothesis testing, ANOVA, regression, ARIMA/Prophet forecasting, linear programming, risk-aware routing, geospatial analysis.
Application development
React, Next.js, Streamlit, responsive web applications, dashboards, document workflows, admin tools, user-facing decision support.
Deployment & systems
Vercel, Railway, cloud PostgreSQL, offline-first design, local/cloud synchronisation, encrypted relay, monitoring, deployment scripts, documentation.
Education & learning
Formal education
and current development.
Credentials already earned are separated from courses currently in progress.
Undergraduate · Expected February 2027
B.Tech Statistics — GPA 4.30/5.00
Federal University of Technology Minna (FUTMinna). Relevant areas include statistical inference, data mining, operations research, linear programming, and multivariate methods.
2024–Present
BSc Applied Science — Data Science
WorldQuant University. Current learning includes the Computer Vision course.
In progress
AI Engineering
DeployEdu — practical AI engineering coursework.
Completed / earned
Selected credentials
MongoDB Certified Developer Associate (2026); WorldQuant University Data Science Lab (2025); ACE Academy Business Management Training & AI for Startups (2026).
Coursera / DeepLearning.AI — MLOps Specialisation In progress
Coursera — Database Management & Design In progress
Research & analysis
Questions explored
through data.
Research activity complements the engineering work, bringing statistical methods and evidence appraisal into applied problem solving.
Completed · Presented
Beyond Capacity — Correctional Overcrowding & Recidivism
Analysed 14 years of NBS/NCoS data across 37 states using forecasting, clustering, and association analysis to examine prison occupancy and recidivism patterns. Presented at DSN AI Bootcamp 2026.
In preparation
Pre-Trial Detention Crisis in Nigerian Correctional Facilities
Longitudinal analysis of awaiting-trial detainee trends across Nigerian states (2011–Q2 2025), including statistical assessment of the ACJIA 2015.
In preparation
AI & Statistics Students: Career Goals, Paths and Perceptions
Multi-university survey research on career satisfaction, AI exposure, and disciplinary identity among Nigerian statistics students.
B.Tech thesis · 2026–2027
Determinants and Demand Projections of Agricultural Mechanisation Adoption
Applies discrete-choice models and demand projection to analyse mechanisation adoption drivers across Nigerian agricultural zones using NCAM field data.
Awards & recognition
Selected milestones.
Recognition is listed as context for the work, not as a substitute for project evidence.
1st Place — System Design & Innovation, HealthTrace Hackathon 2026 (ACTRG/ASBAMI)
3rd Place — NAPE AICE Hackathon 2026
2nd Place — DSN AI Bootcamp, Day 2 (over 120 participants)
Shortlisted 69 of 240 — ESG Student Innovation Challenge 2026
Academia of the Year 2024/2025 Session
Executive of the Year 2024/2025
FUTMinna — Top 5 nomination, AI School of the Year
Contact
Interested in building
something useful?
Open to AI engineering and applied ML opportunities, data and backend engineering roles, research collaboration, and product development work across health, agriculture, education, and civic technology.
Email me ↗