Intelligence engineering · Harare, Zimbabwe
ZJDI designs, builds and secures AI, automation and data systems for Zimbabwean organisations, from the first model to production and beyond, with security as the default.
Avenues Clinic
Ministry of Primary and Secondary Education
NativeHealAvenues ClinicMinistry of Primary and Secondary EducationNativeHeal
Our product · Launching soonSend, sign, track and verify documents in minutes. Face sign-in, signers verified by SMS, a tamper-evident audit trail and a QR code on every signed PDF that anyone can check.
Launching soon. Early-access organisations are onboarded first.
The problem
The ZJDI motto, and the standard we hold every engagement to.
What we do
We will tell you plainly whether AI, automation, better software or a simple process change is the right fix, then build it to run in production.
01 · AI engineering
Machine learning that scores risk, flags fraud, reads documents and recognises faces, deployed as secure APIs your systems can call in real time.
02 · Robotic process automation
Software robots and intelligent workflows that reconcile, notify, capture and file, quietly and accurately, around the systems you already run.
03 · Data engineering
Pipelines, data models and dashboards that give leadership the same accurate numbers, refreshed automatically and ready for decisions.
04 · Web engineering
Customer portals, onboarding flows, admin systems and APIs, engineered for speed on Zimbabwean mobile networks and built to be maintained.
05 · Security & DevSecOps
Security-first architecture, privacy controls, hardened delivery pipelines and reliability engineering for systems that handle sensitive data.
Outcomes
Risk models score every transaction and explain each flag, so analysts work the cases that matter.
Face-verified digital onboarding replaces branch visits and manual document checks.
Bots reconcile, notify and report on schedule, and hand people only the exceptions.
Monitoring, tested backups and clear runbooks are part of every delivery, not an extra.
Selected work
Our product · SaaS01Our own electronic signature platform: send, sign, track and verify documents, with face sign-in, signers verified by SMS and a tamper-evident audit trail.
A QR-launched onboarding platform that verifies customers with deep-learning face matching and ID document recognition, now running in production.
Banking03Machine learning that scores transactions for fraud and money-laundering risk and flags suspicious activity with actionable intelligence.
Banking04Production-ready CatBoost and NLP models that predict loan approvals and recommend suitable amounts in real time.
Agriculture05An AI farming assistant: image-based pest and disease detection, soil analysis, market forecasting, activity tracking and farmer education.
E-commerce06Detects online shopping scams by analysing product links, seller credibility and review patterns, then generates a trust report before you buy.
How a project runs
A clear, low-risk path. You see working software early, decide with evidence and never wait months for a surprise.
01
We map the process with the people who run it, and measure where time, money and accuracy are lost.
02
Architecture, risks, integration points and expected return, agreed with you before any code is written.
03
A working version on your real data, early, demonstrated every two weeks so you can steer.
04
Security review, testing, rollout, training and documented handover to your team.
05
Monitoring, support and continuous optimisation, including model retraining for AI systems.
Industries
Fraud and AML detection, credit decisioning and face-verified digital onboarding, built around core banking systems.
Systems running in production
Help-desk and ticketing, workflow automation and service dashboards that cut delays across clinics and hospitals.
Trusted by Avenues Clinic
Computer vision for pest and disease detection, forecasting and farm intelligence built for field realities.
Predictive maintenance, production reporting and safety analytics for operations where downtime is expensive.
Scam and trust detection, fast storefronts, stock and order automation, and analytics that protect revenue.
Digitised services, case workflows, attendance verification and transparent dashboards for public institutions.
Trusted by the Ministry of Primary and Secondary Education
“The automation suite freed our operations team from manual reporting and doubled our productivity.”
Insights
WhitepaperBanking & AML
Why rule-based fraud monitoring falls behind, and how interpretable machine learning scores transactions in milliseconds without becoming a black box.
Webinar guideAutomation
The companion guide to our recorded session on finding automation opportunities, building resilient bots and keeping them secure.
Best-practice guideSecurity & governance
Practical standards for protecting data when you introduce AI and automation: zero trust, encryption, masking and keeping AI on your own infrastructure.
FAQ
If yours is not here, ask us. A real engineer will answer.
It depends on scope, integrations and data. Every engagement starts with a free assessment, after which we give you a fixed, itemised estimate, so there are no surprises.
We work in two-week sprints and show working software at the end of each one. Many teams see a first working prototype on their own data within the first few sprints.
No. Assessing and preparing data is part of our work. We will tell you honestly what your data can support today and what it would take to do more.
Yes. Much of our work integrates with older core systems and databases such as IBM DB2. Where there is no modern API, we build a secure integration layer around them.
Encryption in transit and at rest, least-privilege access, masking of personal data in analytics and AI, and the option to keep everything on your own infrastructure.
Yes. We are based in Harare and work with organisations across Zimbabwe and the region, remotely and on site as the project requires.
Yes. We offer ongoing support with agreed response times, monitoring and, for AI systems, scheduled model reviews and retraining.
Our goal is to remove repetitive, error-prone work, not people. In practice teams move on to exception handling, customer work and analysis that was always being squeezed out.
Free assessment
An engineer, not a salesperson, will reply within one business day. The first assessment is free and there is no obligation.
