Crop Doctor
Identifies maize and bean leaf diseases from a phone photo and suggests treatment.
We design and build AI, cybersecurity, IoT and software products for businesses across East Africa. Explore our work, then get an estimate in minutes. Browse what’s live, follow what’s in progress, and add your own.
Pick a track to see how we can help and the work we’ve shipped.
Models and assistants trained on local data and languages.
Tools that detect, audit and defend real systems.
Web, mobile and device software built to be maintained.
Sensors, embedded firmware and edge intelligence for the real world.
A clear process with milestones you can see and approve.
We map your goals, users and constraints, then agree scope and timeline.
A clickable prototype or hardware proof of concept, so you validate early.
Weekly demos in short sprints, with security testing built in.
Deployment, training, monitoring and ongoing maintenance.
Mobile money, offline-first and low-bandwidth by design.
Strong authentication, encryption and audit logs on every build.
One team for devices, backend, AI and dashboards.
Full source code, documentation and handover.
Starter patterns from each track. Copy them into your own projects.
#include <WiFi.h>
#include <PubSubClient.h>
void loop() {
float amps = readCT(34); // transformer current
float temp = readTemp(35); // winding temperature, C
if (temp > 85.0) publish("grid/t014/alert", "OVERHEAT");
publish("grid/t014/data", json(amps, temp));
delay(5000);
}from tflite_runtime.interpreter import Interpreter
model = Interpreter("leaf_disease.tflite")
model.allocate_tensors()
def predict(image):
model.set_tensor(INPUT, preprocess(image))
model.invoke()
return LABELS[model.get_tensor(OUTPUT).argmax()]from collections import Counter
def find_bruteforce(events, limit=5):
fails = Counter(e.ip for e in events if e.failed)
return [ip for ip, n in fails.items() if n >= limit]Rings are scored automatically from maturity, adoption and impact. Watch the tour, filter by track, or score your own trend.
Answer six quick questions to see where your business stands. Nothing is sent or stored.
Identifies maize and bean leaf diseases from a phone photo and suggests treatment.
Translates between English, Luganda and Swahili for chat and customer support.
Turns long visit notes into short, structured summaries for clinicians.
Flags phishing links and fake mobile money messages before users tap them.
Reads server logs, spots unusual logins and sends alerts to your phone.
Audits passwords and access policies against known breaches and weak patterns.
A Laravel package for mobile money payments with webhooks and retries.
CI/CD templates that test, build and deploy Laravel and Vue apps in minutes.
Generates searchable API docs from your routes and keeps them in sync.
ESP32 sensors track load, temperature and oil level on distribution transformers and alert before they overload.
Logs temperature and door events for vaccine and produce storage, with SMS alerts when limits are crossed.
Shows output, battery state and faults for off-grid solar sites on one dashboard.
Runs small vision models on-device to count vehicles and detect faults without sending video to the cloud.
A live simulation of a feeder and its transformers for testing load growth before installing hardware.
Choose what you need and see an indicative budget and timeline instantly. We refine it together after a short discovery call.
Share a few details and we’ll reply with next steps and a proposal.