Oluwasola OyinloyeGeospatial data scientist
Geospatial AI specialist
I do GIS, remote sensing and earth observation, drone mapping and geospatial AI work for plantation operators, agric lenders, exporters, developers, governments and NGOs, in Nigeria and abroad. I turn imagery and location data into answers, and I tell you how far to trust each one.
Selected work
What I built and how far it has been tested.
For a commercial oil palm plantation in West Africa, a YOLOv8 model finds each palm crown in aerial imagery and turns it into a mapped, countable tree, block by block. On held-out test tiles it counted 519 palms where a hand count found 518.
302,412 double counts removed (29.6%). Every hectare is attributed to exactly one block.
A machine-learning model that maps which parts of Amuwo-Odofin are more prone to flooding, built on satellite data, published as a live dashboard and written up as a preprint.
The spatial score tests the model on areas it never saw in training. The drop means it can point to areas worth checking, but it shouldn't decide anything about a particular site yet.
Built on Ookla's mobile speed tests, a 1 km map of what every part of Nigeria needs for broadband, and a choice of which clinics and schools should host new sites so that each one reaches as many unconnected people as possible.
How complete is OpenStreetMap's record of health facilities in Lagos? The first phase audits it ward by ward against a reference list of known facilities.
Counts alone can't show completeness. A ward with few clinics on the map may simply have few clinics.
What OpenStreetMap holds today: 605 health facilities across Lagos State's 20 LGAs, pulled 2 October 2026. Whether that is most of them or a fraction is what the audit will measure. Map data © OpenStreetMap contributors.
| LGA | Hospital | Clinic or doctor | Pharmacy | Other | Total |
|---|---|---|---|---|---|
| Alimosho | 181 | 2 | 5 | 0 | 188 |
| Mushin | 103 | 0 | 0 | 0 | 103 |
| Amuwo Odofin | 90 | 0 | 0 | 0 | 90 |
| Ojo | 82 | 0 | 1 | 2 | 85 |
| Eti Osa | 17 | 3 | 14 | 2 | 36 |
| Kosofe | 13 | 4 | 7 | 0 | 24 |
| Ìkòròdú | 4 | 5 | 6 | 0 | 15 |
| Shomolu | 4 | 3 | 2 | 5 | 14 |
| Lagos Island | 5 | 7 | 0 | 0 | 12 |
| Surulere | 7 | 0 | 1 | 0 | 8 |
| Oshodi/Isolo | 2 | 3 | 2 | 0 | 7 |
| Badagry | 4 | 0 | 0 | 2 | 6 |
| Ikeja | 2 | 1 | 3 | 0 | 6 |
| Lagos Mainland | 2 | 1 | 3 | 0 | 6 |
| Ajeromi/Ifelodun | 3 | 1 | 0 | 0 | 4 |
| Ifako/Ijaye | 0 | 0 | 0 | 1 | 1 |
| Agege | 0 | 0 | 0 | 0 | 0 |
| Apapa | 0 | 0 | 0 | 0 | 0 |
| Epe | 0 | 0 | 0 | 0 | 0 |
| Ibeju Lekki | 0 | 0 | 0 | 0 | 0 |
Products
Free public tools I build and maintain alongside client work. Each one documents its data sources and known limits.
Where Amuwo-Odofin is structurally prone to flooding, whether rain is activating that risk now, and which people, schools, clinics and roads are exposed.
ForPlanners, emergency responders, NGOs and researchers
The first phase measures how complete OpenStreetMap's record of health facilities is, ward by ward, so that any map of access to care starts from checked data.
ForHealth planners, NGOs and the OpenStreetMap community
I build open products where good data can help many people at once. If you hold data, fund public-good work or have a question worth answering in the open, let's talk.
Who I work with
How many palms do we have, block by block, and where are the gaps?
Tree counts and stand maps from drone surveys planned and flown in-house, or from satellite imagery.
Is this farm real, how big is it, and is it planted?
Farm boundary, area and crop checks from satellite imagery, before and after you lend.
Can we show EU buyers that our plots weren't deforested after 2020?
Plot geolocation checks and deforestation screening for the EU Deforestation Regulation.
How prone to flooding is this site, and what sits around it?
Flood susceptibility and exposure screening before you buy, design or build.
Where are people and assets exposed, and where should services go next?
Flood exposure, access to services and land-cover change, with the data and code your team keeps.
Who is far from a clinic, and can we trust the map data?
Access and coverage maps, and audits of the data your programme relies on.
Nigeria
On site
Drone surveys planned and flown in-house, GNSS ground control and field checks, through to analysis and delivery.
Across Africa
Remote, satellite first
Open satellite data covers the whole continent, so work on land cover, vegetation, flood risk or change over time can start without a site visit.
Worldwide
Remote, on your schedule
We agree the scope in writing, I send regular updates, and you get standard formats (GeoTIFF, GeoPackage, PostGIS) with the code. Lagos time (UTC+1) overlaps most of the European working day and the morning in the Americas.
Services
10 service areas, from drone surveys and site selection to EUDR checks and training, for clients in any industry. Choose yours to see where I can help.
Deep-learning models that find and count objects in imagery, and predictive models tested on ground they have not seen.
How it worksSurvey flights planned and flown in-house, then turned into orthomosaics, surface and terrain models, and counts.
How it worksLand cover, vegetation, water and change over time, read from optical and radar satellite imagery, anywhere.
How it worksPlot points and polygons checked and screened for deforestation since 31 December 2020, ready for your EU due diligence.
How it worksWhere to build, open, expand or invest: candidate sites scored on access, population, competition, risk and cost.
How it worksSite selection, access and proximity analysis, map production and clean, well-structured spatial data.
How it worksWhere flooding is more likely, which people, roads and facilities sit in the way, and how land cover is changing.
How it worksInteractive web maps, Streamlit apps and Power BI dashboards that work on any device.
How it worksPostGIS databases and repeatable Python and Earth Engine pipelines that keep spatial data current.
How it worksHands-on Python, GIS and remote sensing training for teams, from small workshops to seminars that have reached 1,500 people.
How it worksAlso: cartography and map production for reports, digitising and data cleaning, geocoding, OpenStreetMap mapping, site suitability studies and GNSS ground control. If your question involves a place, ask.
Process
A 15- or 30-minute call: you tell me the decision you need to make, and I tell you whether spatial data can help and what it would take.
I test the approach on a small part of your area and report its error rate, so you can judge it before you commit.
You get the results, the code that produced them and documentation your team can follow.
Engagements
Pick the shape that fits your budget and how sure you are of the approach.
A fixed-scope test on part of your area. You see the method, the output and its error rate before you commit to the full job.
A defined deliverable, such as a tree inventory, a flood screening or a web map, against a written scope, timeline and acceptance criteria.
The same analysis repeated monthly, by season or by year, for programmes and portfolios that need to track change.
Strategic and technical advice by the day or month: what to build or buy, a review of your data and methods, or a geospatial specialist for your bid.
Workshops built around your team's own data, in person in Nigeria or online.
Every engagement starts with a written scope. NDA on request. How I handle your data
About
I'm Oluwasola Oyinloye, a geospatial data scientist and geospatial AI specialist based in Lagos.
I trained in Surveying and Geoinformatics at the University of Ilorin (BSc), then earned a Master of Information Technology (MIT) with an AI specialisation from Miva Open University.
I have five years of professional GIS and remote sensing work behind me, and I have taught Python for geospatial analysis to 1,500 people through seminars.
I take on drone surveys in Nigeria, planned and flown in-house, and work remotely from satellite data for clients across Africa and beyond.
I report how every result was validated, including when the numbers are less flattering. Alongside client work, I build free public tools like the Lagos Access Map and GeoAID, which is written up as a preprint.
Contact
Tell me what you need to decide, where, and what data you already have. I'll tell you plainly whether I can help. Book 15 minutes to introduce yourself or 30 to scope a project, at a time shown in your own time zone. If you'd rather write, the email opens with a short template.
You'll speak with me directly.Lagos, Nigeria · WAT, UTC+1
