The business behind the work
AfterSchoolAfrica helps African learners and professionals discover scholarships, fellowships and development opportunities. Its publishing operation grew to involve more than 15 writers alongside social-media and editorial coordination roles.
As the economics of digital publishing changed, the business needed to preserve useful coverage with a much smaller team. The pressure extended well beyond writing articles: finding opportunities, checking official sources, identifying recurring programmes, tracking changing details and keeping a large WordPress archive useful.
A transparent relationship
AfterSchoolAfrica was founded by Ikenna Odinaka Chukwujekwu, who also founded PrimalIntell. This case study describes operating work within a founder-connected business. It is an internal experience account, not an independent client endorsement.
The problem: recurring work depended on too much manual checking
Every opportunity carried decisions that people had to make repeatedly. Was this genuinely new, or another edition of a programme already covered? Was the deadline still current? Did a funding figure describe this offer or an entire programme portfolio? What needed an editor’s attention?
- Duplicate and recurring opportunities could appear as unrelated posts.
- Deadlines, eligibility and funding details could become stale or be inferred incorrectly.
- Editorial knowledge was difficult to preserve consistently across a large archive.
- Routine checking competed with editorial judgement and more valuable work.
- Uncontrolled AI publishing could introduce unsupported claims.
First, make the operating rules explicit
The work began by documenting how the operation should make decisions. Missing facts remained missing. Uncertain dates did not become invented deadlines. Repeat editions were recognised consistently. People retained responsibility for publication.
These rules became the foundation for automation, so reliability did not depend on each person remembering every exception.
Process and evidence rules
Conservative policies govern identity, deadlines, amounts, eligibility, geography and review.
Deterministic checks
Fixed rules identify duplicates, stabilise opportunity identities, set status and validate outputs before the next step.
Connected workflows
WordPress queues the work; n8n coordinates source selection, extraction, AI support and structured updates outside live page requests.
Human governance
New and changed content is prepared for editorial review. Uncertainty and exceptions remain visible to people.
Then, let automation carry the repetition
AI supports evidence-bound extraction and drafting. Rule-based gates handle identity, routing and validation. Blocked, review and retry states make failures visible instead of silently passing incomplete work through the system.
Editors can direct their attention toward source quality, judgement calls, exceptions and final publication decisions. The system supports their decisions while preserving accountability.
What changed in practice
- More than 300 duplicate URLs were consolidated with canonical signals during the archive clean-up.
- More than 400 obsolete posts were deliberately retired with HTTP 410 responses.
- Opportunities can be classified as open, upcoming, closed or unknown without guessing missing years.
- Stable identities help distinguish genuinely new opportunities from recurring editions and updates.
- Documented workflows and reliable systems support a two-person operating team in managing a core workload previously distributed across a 15-person team.
The capacity result describes how the work became manageable with a smaller operating team. It is not a claim that automation eliminated 13 jobs. The important change was freeing human attention for judgement, creativity, communication and more impactful work.
Evidence note: these figures describe documented internal operating experience. They do not establish independently audited financial returns, revenue recovery or a measured percentage reduction in working hours.
What another business can take from this
The strongest opportunities often sit between tasks: the repeated decisions, hand-offs and checks that keep an operation moving. Making one task faster is useful; making the whole process clearer and more dependable can matter more.
- Diagnose before building. Understand the workflow, ownership, bottleneck and intended outcome.
- Use rules where rules are enough. Apply consistent checks before introducing AI judgement.
- Give AI evidence and boundaries. Validate its output and route uncertainty to people.
- Equip the people doing the work. Preserve accountability and make exceptions manageable.
- Measure before expanding. Compare one bounded improvement against a baseline.
These principles inform PrimalIntell’s workflow automation and workforce enablement engagements for businesses in Nigeria and Africa.