Home · Cohorts

Cohorts

Delivery is not a distribution event. It is a cohort of named people: taught, assessed, certified, and followed up. Below is how CJBest runs that for each pillar, and what the enterprise system was built to survive, which is no network and no light.

One cohort model, six applications

Pillar	The cohort	What CJBest runs	What the system captures	What the leader gets
01 Youth empowerment and skills	Trainees, 30 seats per cycle	Curriculum, induction, and assessment against a pass standard the office sets	Attendance at the session, assessment outcome, certificate reference	A certified register, not a headcount
02 GbajaRide and the fleet	Operators and drivers	Operator onboarding, assignment discipline, service routine	Vehicle to operator, hours reported, service events, faults	Utilisation per vehicle, and the idle ones named
03 Women and youth initiative	Cooperative or individual members	Selection, training, savings-group discipline, follow-up	Membership, training completion, follow-up at 3, 6 and 12 months	Sustainment rate, not just a start date
04 Leadership and mentorship	Mentees and mentors	Intake, session cadence, progression review	Session log, mentoring hours, progression notes	An alumni register with evidence of what changed
05 Constituency record	Mobilisers and clerks, the people who hold the register	Training on capture, duplicate review and consent	Sealed records, chain position, verification checks	A register that survives being questioned
06 Delivery evidence	Field agents	Capture discipline, exception follow-up, recovery of the missing	Coordinates, photographs, serials, exceptions	A ward map and a list of what is not real

Completion is always measured against enrolment, never against seats. An empty seat must never read as a success.

Built for the two barriers that actually stop youths

Programmes fail in the field for reasons that have nothing to do with curriculum. There is no signal, and there is no power. The enterprise system was designed around both rather than apologising for them afterwards.

No network

Offline first. Capture works with no connection at all. Every record writes to the device first, queues, and syncs when a link returns. A dropped connection is a delay, never a lost session.
Idempotent sync. The same record sent twice creates one row, not two, so a flaky link cannot corrupt the register.
Cached application. The app shell and the current cohort list live on the handset, so it opens and works with the radio off.
Text first, small payloads. Records are compact and photographs are downscaled on the device before they queue. Verification never requires video.
Paper bridge. Where nobody has a smartphone, the session runs on a printed sheet carrying a sealed code that is scanned in later, so a cohort is never split by who owns a phone.
Proof before connectivity. Each record's fingerprint is computed on the device and chained, so the evidence exists before the network does.

No light

Low-power profile. Dark interface, no animation, no background polling. A session is designed to fit inside a phone battery rather than test it.
Daylight scheduling. Cohorts are planned around daylight. The system never assumes an evening session with mains power.
Charge at the venue. Solar banks sized per cohort and carried as a line item in the programme cost, not an afterthought after the first dead phone.
Resume after interruption. If a phone dies mid-session, queued records survive and the session continues on a second device or on paper. Nothing is repeated and nothing is lost.
Sparse sync. Data moves in the smallest useful batch when a link appears, so a single charge buys more sessions.

What makes it hold beyond the software

A named owner per ward. Someone, by name, holds the register. Systems do not maintain themselves.
Mobilisers are trained and assessed the way trainees are. Capability is certified, not assumed, because the register is only as good as the person capturing it.
The exception list is published to the office monthly, including the parts that did not work.
Follow-up is the product. Three, six and twelve months, or the whole thing is a photograph of a good day.

Start with one cohort

The cheapest way to prove all of this is one ward, one cohort, one cycle: 30 seats, one assessment, one certificate, one follow-up date. That produces the case study which makes the next five cohorts easy, and it can be running inside a month.

Request a scoping session See every programme