From our work
ACBF
Public Sector Collaboration Hub for the African Capacity Building Foundation
Continental
The sequential trap
Continental knowledge platforms have a familiar failure mode. The engineering team builds the platform first, on assumptions about what the content will look like. The research team produces the content second, on a schedule that runs after the engineering decisions are already locked in. By the time the research outputs are ready, the platform's data model, taxonomy, and analytics surface have been engineered for content that does not match what the research actually produced.
The result is a platform full of compromises. Categories that do not match the analytical framework. Performance scales that do not respect the methodological integrity of the underlying assessments. Search behaviour that retrieves the wrong content because the tagging schema was designed before the research team could specify what it should be. Users sense the misalignment immediately, and the platform never quite earns the trust it was built to deserve.
Parallel workstreams as a design discipline
The alternative is to treat engineering and research as parallel workstreams that inform each other in real time, rather than sequential phases that hand off to one another. The discipline requires structural commitments. The data model is co-designed with the research team, not derived from a generic platform template. The analytics surface is built around the harmonisation methodology, not retrofitted to it. The taxonomy and tagging schema emerge from the research workstream and become engineering constraints, not the other way around.
The cost is real. Both teams have to operate at the same pace, with the same level of decision-making authority, on a schedule that requires more coordination than a sequential project plan. The benefit is greater. The platform that emerges actually fits the content it was built for, because the content was the design input rather than the deliverable.
What this looks like at ACBF
PANEOTECH, in joint venture with JAMII LAB, is delivering the Public Sector Collaboration Hub for the African Capacity Building Foundation, a specialised agency of the African Union, with funding from the Bill and Melinda Gates Foundation. The engagement runs both workstreams in parallel from inception. JAMII LAB leads the analytical and research component, including diagnostic harmonisation across PEFA, TADAT, PIMA, SAI PMF, and MAPS, the continental ecosystem mapping, and the case study compendium. PANEOTECH leads the technical implementation, with continuous coordination between the two teams.
The data model reflects the harmonisation methodology rather than constraining it. The analytics surface renders the harmonised performance bands JAMII LAB produces, with the comparison guardrails the methodology requires. The content blueprint, knowledge taxonomy, and case study format originate in the research workstream and shape the platform's editorial workflow rather than being grafted onto it. The platform launches with content the research team designed it to host, not content the research team had to make fit.
The institutional lesson
For continental knowledge platforms with a long horizon, the joint venture between engineering and research is structural, not contractual. Treat it that way and the platform becomes a coherent continental asset. Treat it as a procurement convenience and the platform becomes another tool nobody quite trusts.
We build the platforms that institutions stand behind.
Joint engineering and research workstreams, structured editorial governance, and the institutional discipline that continental clients require.
