ODIT Frontier Partners
The African Investment Firm After Headcount
oditfrontierpartners.net
ODIT Frontier Partners
AI Subsumption Series
AI Subsumption Series · Excerpt 02 of 10 · Part I
Mapping the Jobs, Functions and Structures of African Venture Capital and Private Equity Funds
To establish the baseline before assessing AI subsumption.
Thirty-eight organisations reconstructed, classified and redrawn to identify the canonical structures from which the modern African investment institution is built.
This is Part I of The African Investment Firm After Headcount (Dispatch D-008, AIS-05), a seven-part, twenty-six-table research report from Odit Frontier Partners. That report sits within the wider AI Subsumption Series, opened in D-003 (the displacement thesis) and continued in D-007 (Bypassing the GP).
Before anyone can say what artificial intelligence will subsume inside the investment firm, someone has to decompose what the investment firm actually is. You cannot identify what goes extinct without first decomposing the roles.
This excerpt publishes the first step of that work: the reconstructed organisational structures of 38 African investment institutions, anonymised, grouped by weight class, each drawn twice, as the firm presents itself and as the firm actually functions.
The Mission
Map the organisational structures of 38 African funds, decompose them into functions, compress the functions into unified archetypes, and only then ask:
Which of these seats can artificial intelligence subsume?
38 funds → taxonomy → 7 archetypes → the canonical fund → the subsumption matrix
Part I · Mapping the Institution · The Map
The Examination
What, in the actual architecture of the African investment fund, can artificial intelligence subsume; what can it not; and what institution remains?
AI compresses labour. It does not eliminate institutions.
The Series · Ten Excerpts
Where We Are in the Argument
Everybody talks about AI subsumption; nobody has mapped what is being subsumed.
This excerpt is the anatomy: the sample, the method, all thirty-eight structures, and the two observations they force. The compression analysis that follows from them is the subject of the excerpts ahead.
How to read this
Contents.
This excerpt runs in five blocks, from why the study exists to the matrix that tests it. The 38 reconstructed organograms sit in Part II; the compressed archetypes they resolve into sit in Part IV.
The Problem
Why the study exists: anatomy before extinction.
The Sample
38 firms, randomly drawn, in three weight classes.
The Method
Each firm reconstructed twice: observed, then functional.
The Thirty-Eight Structures
All 38 organograms, grouped by class (MV, MR, UP).
Building the Taxonomy
The failure of titles and the canonical ten functions.
The Seven Archetypes
The composite shapes the 38 resolve into.
Preparing for Subsumption
The minimum viable structure, LP verification, the matrix.
The Close & Back Matter
The unavoidable next question, sources and colophon.
First Principles
What these firms are, and who works in them.
A venture capital, private equity or impact fund is a firm that raises a pool of money from outside investors, chooses a small number of companies to put it into, works to make those companies more valuable, and returns the proceeds, keeping a share of the gains as its fee. The whole business is a handful of decisions made by a small number of people, so who sits in the firm and what they actually do is the institution. Every structure in this dispatch is a variation on four things a fund must do.
The four things a firm must do
Source
Find and win deals, the flow of companies to invest in.
Decide
Judge which deals to back and sign the cheque, the investment decision.
Support
Work with the companies after investing to grow their value.
Run & verify
Keep the books, stay compliant, and report to investors, the operating spine.
In developed markets, typically…
The firm is drawn as a pyramid: a few partners who decide and sign over a broad base of associates and analysts who model, screen and run diligence. That the base is the widest layer is the assumption this study tests against the African record.
Block 1 · The Problem
Why this study exists.
- 01Everybody talks about AI subsumption.
- 02Nobody has mapped what is being subsumed.
- 03Before analysing extinction, one must analyse anatomy.
- 04Therefore OFP reconstructed 38 African investment institutions across venture capital, growth capital, impact and DFI-anchored vehicles, control private equity, family capital and captive vehicles.
Block 2 · The Sample
What was reconstructed.
The 38 firms were not hand-picked. They were drawn by random machine selection from the full population of African investment institutions with a documented, publicly traceable team, a computer-generated random draw across the eligible universe, with no analyst curation of which firms entered the sample. This removes selection bias: the sample is a random cross-section of the industry as it exists on the public record, not a set of firms chosen to prove a point.
Table E1 · Regional anchor of the 38 firms (primary operating hub; several are pan-African)
| Region | Firms |
|---|---|
| Southern Africa | 10 |
| East Africa | 9 |
| West Africa | 8 |
| North Africa | 6 |
| Pan-African platforms | 5 |
Table E3 · Institutional forms present (a firm can carry more than one)
| Form | Members |
|---|---|
| Pure partnership boutique | 7 |
| Institutional seed VC | 9 |
| Pan-African growth VC | 4 |
| DFI-anchored impact vehicle | 7 |
Table E2 · Weight classes by total headcount
| Class | Headcount band | Firms |
|---|---|---|
| Minimum-viable (MV) | 4 to 8 people | 11 |
| Mid-range (MR) | 9 to 15 people | 21 |
| Upper institutional (UP) | 17 to 45+ people | 6 |
The three weight classes, defined
The classes are a headcount instrument, but each behaves as a distinct kind of institution, so each is defined in full before any structure is reconstructed. They are carried through the rest of the dispatch as three separate analytical tracks.
MV · Minimum-viable class · 4–8 people · 11 firms
The smallest configuration in which a fund exists at all as a regulated, fundable entity. Apex and judgment are fused into the same two or three people, who source, decide and sign; there is no dedicated analyst layer, and the operating spine (finance, fund administration, compliance) and the verification function are typically off the payroll, carried by external administrators and auditors. This is the institutional floor, the fund reduced to the seats it cannot legally or commercially do without.
MR · Mid-range class · 9–15 people · 21 firms
The modal African fund and the centre of gravity of the sample. Here the full three-chain geometry appears in complete form: a two-person apex, a delegated deal-leadership layer beneath it, a thin execution band, a platform or portfolio-support seat, and a covenant-driven operating spine that the fee income exists to fund. This is the first class large enough to show internal delegation, and the class where the gap between published titles and actual function is widest.
UP · Upper institutional class · 17–45+ people · 6 firms
The control private-equity houses and multi-fund platforms, the only firms in the sample with genuine internal depth. They run multiple parallel deal teams, dedicated in-house finance, compliance and investor-relations functions, and their own verification layer, and the largest of them scale headcount with cohort or platform delivery rather than with analysis. This is the closest the sample comes to the developed-market pyramid, and the benchmark the smaller classes are measured against.
Block 2 · The Sample, At a Glance
Where the thirty-eight sit.
By regional anchor
By institutional form (a firm can carry more than one)
Counts exceed 38 because a single firm can carry more than one form, a boutique that is also DFI-anchored, for instance.
Block 3 · The Method
Reconstructing the institution.
Step 1
Observed structure
Titles exactly as the firm publishes them.
Step 2
Functional structure
Same people, re-sorted by decision rights.
Step 3
Canonical archetype
Compressed to one of seven shapes.
Five public sources per firm
- Public team pages.
- Professional biographies.
- DFI appraisal and disclosure documents.
- Portfolio and deal announcements.
- Regulatory disclosures.
Each institution reconstructed twice
Observed structure, the titles, exactly as the firm publishes them.
Functional structure, the same people, re-sorted by decision rights: who votes, who signs, who executes, who verifies.
Why two drawings of the same firm
The observed structure records the institution as it presents itself. The functional structure records the institution as it operates, because titles in this sample are local dialects, and grade words such as Partner and Principal frequently mark pay bands rather than jobs. Each view reveals what the other cannot.
The observed chart preserves the evidence: title strings are kept verbatim because the titles are the evidence. The functional chart exposes the anatomy, who votes, who signs, who executes, who verifies, and which seats are lateral functions disguised by line-sounding titles. Where the two drawings diverge, the divergence is itself a finding: half the sample’s “middle” is not middle at all; it is sideways.
Evidence tiers, T1 full dual reconstruction (12 firms); T2 partial reconstruction from documented layers (23 firms); T3 count-level only (3 firms).
Confidence markers, [C] confirmed from the record; [SI] strongly inferred; [T] tentative; [U] unknown. Reconstructed lines are inferences and presented as such.
Anonymisation, firms are coded within weight class; names, people, places and portfolio identifiers are removed; title strings are preserved because the titles are the evidence.
Part II · The Thirty-Eight Structures
The Minimum-Viable Class
4 to 8 people · 11 funds
Judgment partly rented; the spine and verification sit almost entirely off the payroll.
Structural Properties
The smallest credible institutional funds run at four to eight people on payroll. Apex and senior judgment are fused in two to four people, and the partners themselves are the investment committee. Execution is zero to one seats. The spine is outsourced by design, fund administrator, external counsel, auditor and part-time finance. Governance independence is supplied from outside the payroll, through external IC members, venture partners and non-executive boards. Fractional and hybrid seats are common.
Typical Forms
The pure partnership boutique, the family capital vehicle, the tax-incentive-era governance wrapper, the accelerator partnership and the offshore-fund-plus-local-adviser geometry. What unites them is that headcount arithmetic fails: a fund of four to eight on payroll is institutionally a roughly ten-seat organism, with about seven seats sitting on other companies’ payrolls.
What Distinguishes the Class
There is no middle to compress. Its binding constraint was always partner hours, not junior headcount, and the verification apparatus that makes it investable is almost entirely external.
A · Observed Title Structure
{{ C.mv07a }}B · Functional Structure (Decision Rights)
{{ C.mv07b }}Structural Notes
US-style vice-president nomenclature in an African firm: the title vocabulary follows the founder’s corporate-VC lineage, not geography. No finance function is visible, fund administration is outsourced [SI], and senior capability is partly rented through a co-managed strategic platform rather than hired. Five to six people cover a seed fund plus a growth platform: function-combination at every seat.
A · Observed Title Structure
{{ C.mv08a }}B · Functional Structure (Decision Rights)
{{ C.mv08b }}Structural Notes
The extreme externalisation case: the investment committee, the senior judgment bench and the control functions all sit outside the payroll. The firm is a chief executive plus a governance wrapper. Origin matters, Section 12J tax-incentive vehicles, a South African regime whose sunset clause ended in June 2021, required strong governance optics with thin economics, and produced exactly this shape.
A · Observed Title Structure
{{ C.mv09a }}B · Functional Structure (Decision Rights)
{{ C.mv09b }}Structural Notes
Two boundary-dissolving roles: a fractional general partner and an LP-IC-founder hybrid. Employment status, capital provision and governance authority are three different things occupying the same two people. Headcount arithmetic fails here: a “team of 8” contains perhaps 5.5 full-time equivalents and two capital-governance seats.
Minimum-Viable Class · Tier 2 · Partial Reconstructions
Reconstructed from documented layers.
Reconstructed Structure (Documented Layers)
{{ C.mv01 }}Structural Notes
The pure partnership: no middle layer to compress at all; lines confirmed by absence. The partnership itself is the investment committee, and the spine is consumed externally, administration, legal and accounting all sit off the payroll.
Reconstructed Structure (Documented Layers)
{{ C.mv02 }}Structural Notes
One team, three institutional wrappers: the same people operate simultaneously as a family office, a fund manager and a contracted corporate-venture manager, so employment status and mandate diverge per fund. The wrappers, not the headcount, are the structure.
Reconstructed Structure (Documented Layers)
{{ C.mv03 }}Structural Notes
An accelerator partnership that rents programme infrastructure rather than building it. The senior execution seat is titled “Principal Investment Manager”: one title straddling both nomenclature families, evidence of dialect mixing at small scale.
Reconstructed Structure (Documented Layers)
{{ C.mv04 }}Structural Notes
A six-person fund whose entire geographic middle layer is fractional: country coverage is rented from venture partners rather than employed. The payroll core is a managing partner and a senior associate.
Reconstructed Structure (Documented Layers)
{{ C.mv05 }}Structural Notes
The maximum-combination case in the sample: one seat holding what larger firms split into five. The founder combines origination, committee authority, LP relationships and eight portfolio board seats in one person, single-GP growth equity at its most concentrated.
Reconstructed Structure (Documented Layers)
{{ C.mv06 }}Structural Notes
A family capital vehicle: the investment committee and the LP collapse into the family, and the fund operates as an execution arm of a balance sheet. Fundraising does not exist as a function.
Reconstructed Structure (Documented Layers)
{{ C.mv10 }}Structural Notes
A mezzanine evergreen in which part of the finance function is a product delivered into investees, not only a control function: designated portfolio accounting staff provide technical assistance inside the small businesses the fund finances. Country investment managers carry the geographic middle; function elevates the title.
Reconstructed Structure (Documented Layers)
{{ C.mv11 }}Structural Notes
Fund entities offshore, advisers on the ground: adviser-model geometry again, with country advisory companies run by local managers as separate legal entities. Also the consolidation vehicle for a legacy East African fund, one of the sample’s two spine-sharing consolidation cases.
Part II · The Thirty-Eight Structures
The Mid-Range Class
9 to 15 people · 21 funds
The modal African fund: a delegated deal layer, a platform function and an owned spine over the minimum core.
Structural Properties
Twenty-one of the thirty-eight funds, the modal form, sit at nine to fifteen people. Above the minimum core the class adds three things: a delegated deal layer (a Principal, Head of Investments or Investment Director leading transactions under the partner group, with investment managers and associates executing), a platform function (founder services, networks, talent, communications), and an owned fiduciary spine, in-house finance ownership over partly external execution, with the finance head managing the external administrator. The spine runs at 30 to 46 percent of headcount.
Typical Forms
Combination seats are the class signature: one person as executive assistant and investor relations, a CFO who also owns LP relationships, a chief of staff carrying investor relations, reporting and operations in a single seat. Where DFI capital is present, the LP register installs ESG and compliance seats regardless of assets under management. Coverage models are emerging as a third chain beside deals and operations.
What Distinguishes the Class
The absence of the structure the textbook predicts: no fund in the sample, of any size, carries a broad-based analyst pyramid, and only roughly one firm in four retains even a single dedicated analyst. The class grew downward from partner clusters, partners first, structure later, inverting the US pattern.
A · Observed Title Structure
{{ C.mr01a }}B · Functional Structure (Decision Rights)
{{ C.mr01b }}Structural Notes
The two-analyst layer exists but is exactly two heads: a capability, not a pool. The Director, Portfolio & Strategy seat is a lateral function disguised by a line-sounding title, portfolio management extracted from the deal chain into its own seat. Finance is held in-house as a Controller and Accountant pair, with the operations director at the firm’s LP-facing node, suggesting fund administration is at least partly in-house, unusual in the sample [T].
Full dual reconstruction on file; detailed chart not reproduced in this excerpt.
A · Observed Title Structure
{{ C.mr03a }}B · Functional Structure (Decision Rights)
{{ C.mr03b }}Structural Notes
Two funds run off one team. The asset-manager portfolio model assigns each professional ten to fifteen companies, formalising portfolio delivery as its own chain. A fractional venture partner and the external board form the governance scaffolding above the payroll.
Group structure, three legal entities: the fund is legally a regulated manager plus two advisory companies.
A · Observed Title Structure
{{ C.mr04a }}B · Functional Structure (Decision Rights)
{{ C.mr04b }}Structural Notes
The organisation chart is legally three companies: a regulated manager and two African advisory companies. The African teams advise; the regulated entity decides, as a regulatory matter, so decision rights and the visible org chart diverge by construction [C]. “Associate Investment Director” is a genuine intermediate rung, seen only here: evidence that ladders grow rungs organically.
A · Observed Title Structure
{{ C.mr05a }}B · Functional Structure (Decision Rights)
{{ C.mr05b }}Structural Notes
The only fund in the sample with the full hypothesised East African deal ladder visible on one page: Investment Director, Senior Investment Manager, Investment Manager, Investment Associate. Evergreen family capital changes the apex, the family board sits above a professional CEO, and in-house legal is a rarity in the sample.
A · Observed Title Structure
{{ C.mr06a }}B · Functional Structure (Decision Rights)
{{ C.mr06b }}Structural Notes
A dual-fund, DFI-heavy LP base drives an unusually formalised control layer for a firm of this size: a dedicated compliance and AML head plus a partner-grade ESG seat. The compliance floor is LP-determined, not size-determined. The promotion record shows the firm began as a pure partner cluster and grew a structure downward.
A · Observed Title Structure
{{ C.mr07a }}B · Functional Structure (Decision Rights)
{{ C.mr07b }}Structural Notes
The single most important structural datapoint in the sample for the subsumption thesis: the firm achieves capability without central headcount by exporting its middle layer into portfolio companies. The centre is three people; the institution is nine; the capability footprint is the whole portfolio. It is human-capital arbitrage doing the job the AI thesis assigns to software, proof that the geometry works, independent of the technology used to achieve it.
A · Observed Title Structure
{{ C.mr08a }}B · Functional Structure (Decision Rights)
{{ C.mr08b }}Structural Notes
Direct documentary evidence of the outsourced fiduciary model: the CFO’s published role is to interface with an external fund administrator, in-house ownership over external execution. The Chief of Staff is a three-function seat combining investor relations, reporting and operations, the growth-fund equivalent of the EA/IR combination seen at seed scale. “Founding Partner” here means originating sponsor, not operating executive, a fourth meaning of “Partner” in the sample.
Mid-Range Class · Tier 2 · Partial Reconstructions
Reconstructed from documented layers.
Reconstructed Structure (Documented Layers)
{{ C.mr09 }}Structural Notes
Classic three-rung venture structure (general partner, principal, associate) with talent as the platform specialisation. A verified promotion chain runs associate to principal to general partner, with board responsibilities beginning at the principal rung: the clearest documented career ladder in the sample.
Reconstructed Structure (Documented Layers)
{{ C.mr10 }}Structural Notes
A studio-plus-fund hybrid: the studio builds companies, the fund invests, and staff span both. Operating partners are full-time portfolio commercial operators, not deal staff. A DFI appraisal gives unusually good functional detail, including named ESG and compliance seats, the DFI floor made visible. The legal wrapper sits in Europe with the operating team on the continent: adviser-model geometry again.
Reconstructed Structure (Documented Layers)
{{ C.mr11 }}Structural Notes
Distributed pods rather than a headquarters pyramid: investment and operations personnel sit across the major African technology markets. Partner-dense, with several venture partners likely inside the published count, an upper-bound headcount, not a payroll [T].
Reconstructed Structure (Documented Layers)
{{ C.mr12 }}Structural Notes
External venture partners holding formal investment-committee votes: rented judgment with real decision rights, a stronger form than the advisory venture partners seen elsewhere. The partner-grade CFO owns finance and LP relationships in one seat, a control function combined with investor stewardship.
Reconstructed Structure (Documented Layers)
{{ C.mr13 }}Structural Notes
A governance function unusually senior for the firm’s size, and deal leadership sitting at partner level [C from deal record]. The analyst seat was added only recently: junior analytical capacity arrives late, if at all, in this class.
Reconstructed Structure (Documented Layers)
{{ C.mr14 }}Structural Notes
An on-balance-sheet evergreen: no LP fundraising cycle, therefore no investor-relations seat anywhere in the structure. Capital structure deletes an entire function. The partner seat carries documented promotion evidence from principal dealmaker, another partners-first ladder grown downward.
Reconstructed Structure (Documented Layers)
{{ C.mr15 }}Structural Notes
A corporate evergreen: the LP function is replaced by a parent-group relationship, and value creation is a formalised bridge between the portfolio and the owner’s operating companies. Finance and operations share one combined seat.
Reconstructed Structure (Documented Layers)
{{ C.mr16 }}Structural Notes
A two-MP apex with a senior operations partner sitting laterally beside the deal chain. The gender-lens DFI mandate carries an impact-measurement function [SI]: the LP register, not the assets under management, sets the seat.
Reconstructed Structure (Documented Layers)
{{ C.mr17 }}Structural Notes
Collective and rolling-fund geometry rather than a firm pyramid: the community is the middle layer. Platform capacity is a network, not a department.
Reconstructed Structure (Documented Layers)
{{ C.mr18 }}Structural Notes
Advisory and capital-raising services run alongside the funds: fee income subsidises the spine, another small-fund survival strategy. The deal chain is investment managers under a three-partner apex.
Reconstructed Structure (Documented Layers)
{{ C.mr19 }}Structural Notes
A partnership with a named principal layer and a network standing in for platform staff. As at other funds in this class, the platform function exists; the platform payroll does not.
Tier 3 · Count-Level Reconstructions
Reconstructed at count level only.
| Fund | Status |
|---|---|
| MR-20 | Legacy East African PE manager; absorbed into a larger platform in 2023 specifically to share an operating spine; structure reconstructed pre-absorption at count level. |
| MR-21 | Legacy East African VC manager; management assumed by another regional firm in 2020; the second spine-sharing consolidation case in the sample. |
Part II · The Thirty-Eight Structures
The Upper Institutional Class
17 to 45+ people · 6 funds
Verification internalised, in-house legal, multi-seat finance, formal risk; the operating apparatus is the majority of the firm.
Structural Properties
Six funds sit at seventeen to forty-five-plus people, and the change at this scale is not more deal staff, it is verification brought in-house. Control transactions demand internal legal and finance depth, so the class carries an in-house general counsel (rare everywhere else), a multi-seat finance function, and, at the largest firm, a named Chief Risk Officer: the control layer specialising into finance, risk, sustainability and investor relations as four separate senior seats.
Typical Forms
Three forms. The control private equity house, where the operating spine is the outright majority of the institution, ten of seventeen heads at the exemplar sit outside the deal chain. The platform-embedded team, whose visible leanness is partly an artefact of consuming a shared global spine. And the programme factory, whose headcount scales with cohort delivery, not with analysis: different institutional physics from every other fund form.
What Distinguishes the Class
Where its mass sits. Its weight is accountability-dense, not workflow-dense: the seats that make it heavy exist so that outside parties can rely on the firm, which is precisely why they resist deletion.
Full dual reconstruction on file, the control-PE exemplar, where ten of seventeen heads sit outside the deal chain; detailed chart not reproduced in this excerpt.
Reconstructed Structure (Documented Layers)
{{ C.up02 }}Structural Notes
The only fund in the sample with a named Chief Risk Officer: at this scale the control layer specialises into four separate senior seats, finance, risk, sustainability and investor relations. Principals originate, execute and support portfolio across a multi-office network: the deal rung carries full-cycle responsibility.
Reconstructed Structure (Documented Layers)
{{ C.up03 }}Structural Notes
The self-spine model: at scale the firm internalised what small funds outsource, then sold spare capacity as advisory, “a full investment house” [C]. Also the spine consolidator that absorbed a smaller manager: the sample’s clearest evidence that sub-scale managers merge to share an operating spine.
Reconstructed Structure (Documented Layers)
{{ C.up04 }}Structural Notes
Headcount scales with cohort delivery, not with analysis: an education-and-deployment factory with funds attached, running repeated startup cohorts per office per year. Different institutional physics from every other fund form in the sample.
Reconstructed Structure (Documented Layers)
{{ C.up06 }}Structural Notes
The Africa team’s visible leanness is partly an artefact of consuming a shared global spine: legal, finance, compliance and investor relations are platform services, with a dedicated platform-level partner carrying investor relations. Not comparable with standalone firms without this flag.
Tier 3 · Count-Level Reconstruction
| Fund | Status |
|---|---|
| UP-05 | Infrastructure investment platform; counts only at reconstruction date; included for weight-class and spine observations. |
Block 5 · First Observation
Thirty-eight structures. The structures appeared diverse.
The functions did not.
Across every chart just shown, the same small set of functions recurs under wildly different names. The same title means different things at different firms; different titles mean the same thing. Before the anatomy could be compared, the titles had to be translated. That is the dialect problem, and it is the first finding of the study.
The Finding, In One Picture
The pyramid, inverted.
The developed-market firm is a pyramid, a few partners over a broad analyst base. Across the 38 African firms reconstructed here that shape is turned on its head: the seniors are the wide band and dedicated analysts, where they exist at all, are the thin sliver.
Developed-market model
Junior analytical labour is the widest layer, the base the pyramid rests on.
functions
inverted
The African picture · this study
The dedicated analyst layer is the thin sliver, the point the funnel narrows to.
The inversion is the whole argument in one image: where the developed-market firm concentrates its people in the workflow-dense base, exactly the layer AI is most able to subsume, the African firm concentrates its people in the accountability-dense apex, the layer that resists subsumption.
Part III · Building the Taxonomy
The failure of titles.
Table E4 · The same title, different institutions
| Title | At one firm | At another |
|---|---|---|
| Principal | Leads deals and holds portfolio board seats | A pay grade, applied to the head of platform, the general counsel or the chief financial officer |
| Partner | An investment committee voter and equity holder | A compensation band carried by lateral operating leadership |
| Investment Director | Delegated deal leadership below the partner group | Senior execution within a country office, without committee rights |
One title, several jobs
Left of the divider: what the title claims. Right: what it can also mean in the same sample. The title alone cannot tell you which.
Titles were discarded. Decision rights were retained. The only reliable discriminators across all 38 firms were committee membership, signature authority and board seats.
Part III · Building the Taxonomy
The canonical taxonomy.
Ten functions, into which every observed title in the sample resolves.
Investment chain
Portfolio & platform
Fiduciary spine
The ten functions resolve into the three chains that branch beneath every apex, the geometry the archetypes are built from.
| # | Function | What it holds |
|---|---|---|
| 1 | Apex authority | Committee chair, LP or owner trust, the firm itself |
| 2 | Senior investment authority | Committee votes, capital allocation |
| 3 | Delegated deal leadership | Leads transactions, holds board seats |
| 4 | Execution | Screening, models, memoranda, data rooms |
| 5 | Portfolio | Cross-portfolio monitoring and value creation |
| 6 | Platform | Founder services, networks, talent, communications |
| 7 | Finance | Fund accounting, valuations, administrator relationship |
| 8 | Legal | Documentation and counsel, in-house or rented |
| 9 | Compliance | Named officerships, the covenant calendar |
| 10 | Governance | Independent committee seats, the external verification shell |
Thirty-eight firms entered the reconstruction. Hundreds of titles entered the reconstruction. A much smaller number of functions emerged.
Part IV · The Seven Archetypes
The pure partnership boutique
In sample, MV-01 · MV-02 · MV-03 · MV-04 · MV-05 · MV-06 · MR-17.
Synthesised geometry; archetypes are forms, not partitions, so a firm can carry more than one.
Part IV · The Seven Archetypes
The institutional seed VC
Three short chains sharing one apex: investment, platform, fiduciary.
In sample, MR-02 · MR-03 · MR-06 · MR-09 · MR-10 · MR-11 · MR-12 · MR-18 · MR-19.
Synthesised geometry; archetypes are forms, not partitions.
Part IV · The Seven Archetypes
The pan-African growth VC
In sample, MR-01 · MR-04 · MR-08 · UP-06 (platform-embedded variant).
Synthesised geometry; archetypes are forms, not partitions.
Part IV · The Seven Archetypes
The DFI-anchored impact vehicle
The DFI floor: seats installed by covenant, present regardless of size.
In sample, MR-04 · MR-05 · MR-06 · MR-10 · MR-16 · MV-10 · MV-11.
Synthesised geometry; archetypes are forms, not partitions.
Part IV · The Seven Archetypes
The control PE house
The majority of the institution sits outside the deal chain.
In sample, UP-01 · UP-02 · UP-03 · MR-13 · MR-14 · MR-20 (legacy).
Synthesised geometry; archetypes are forms, not partitions.
Part IV · The Seven Archetypes
The inverted, deployed-middle model
In sample, MR-07 (pure form) · MV-10 · MR-10 (partial) · UP-04 (industrialised).
Synthesised geometry; archetypes are forms, not partitions.
Part IV · The Seven Archetypes
The balance-sheet and captive vehicle
No fundraising cycle, therefore no investor-relations seat anywhere in the structure.
In sample, MR-14 · MR-15 · MV-06 · MR-05 (evergreen) · MV-09 (partial).
Synthesised geometry; archetypes are forms, not partitions.
Part IV · The Canonical African Fund
The composite: one apex, three short chains.
Above the payroll, external governance shell
LP base or family board or parent group; non-executive and external committee members; the fund administrator, the auditor, external counsel.
Chain 1 · Investments
Partners and GPs, the committee voters
Principal / Investment Director / Head of Investments
Investment Manager / (Senior) Associate
Analyst, present at ~1 firm in 4, one to two heads
Chain 2 · Portfolio & Platform
Head of Platform, Director Portfolio & Strategy, Head of Talent.
Chain 3 · Finance, Ops & Compliance
CFO or Controller, fund operations, compliance, EA and administration.
In every fully verified fund, the hierarchy branches directly beneath the apex into three chains: investments; portfolio and platform; finance, operations and compliance. Nothing in the sample shows finance or platform reporting through the investment chain. The single vertical pyramid of the textbook does not exist here.
Part V · Preparing for Subsumption
What exists, stated as a minimum.
The Minimum Viable Institutional Structure (MVIS)
Read as a floor, not a pyramid: five layers that must be present for the fund to exist as a fundable, regulated institution, two of them, by design, sitting off the payroll.
Dashed layers sit off the payroll by design, rented, not employed.
Four structures now stand established from the evidence: the minimum viable institutional structure above; the operating spine, one third to one half of every payroll; the external verification shell, off the payroll by design; and the DFI floor, the additional seats that development-finance covenants install regardless of workload. Now, and only now, the institution is mapped well enough to be tested.
Part V · Preparing for Subsumption
What an LP must be able to verify.
The minimum is not an aesthetic preference; it is what a limited partner minimally requires before a commitment can be considered, and every requirement is stated the way LP covenants state it, as a function with an accountable owner, never as headcount. Each is independently verifiable, and each weight class satisfies it differently.
| What an LP minimally requires | Minimum-Viable | Mid-Range | Upper Institutional |
|---|---|---|---|
| Authorised manager and accountable officers Verified from the licence / register and named officerships |
The partners themselves hold it; the licence and the named officers are the institution. | Held at the apex, with named finance and compliance owners beneath it. | Held across a specialised officer layer: finance, risk, compliance and legal are separate named seats. |
| Independent fund administration Verified from the administrator appointment |
External by design; the administrator is the fund’s accounting department. | Execution is external under an in-house owner who manages the administrator relationship. | Partly internalised, with in-house fund operations alongside the external administrator. |
| Segregation of investment and valuation Verified by NAV produced or checked outside the deal team |
Achieved by geometry: the administrator and the auditor sit outside the firm entirely. | Achieved by the segregation triangle: in-house finance ownership, external administration, external audit. | Achieved both internally and externally: dedicated finance seats plus the external triangle. |
| Named compliance and AML arrangements Verified from the named officer or external provider |
Externally provided, or waived at micro scale, until DFI capital enters. | Move in-house once DFI capital is present; the LP register, not the AUM, installs the seat. | Standing in-house capability, with formal risk added at the largest firms. |
| One IC seat independent of daily execution Verified from the committee composition |
Supplied by external IC members, venture partners and non-executive boards. | Supplied by external members and, in adviser-model funds, by the regulated entity itself. | Supplied by formal governance: non-executive chairs and independent committee seats. |
Part V · The Subsumption Matrix
How subsumption was matrixed.
Every function of every archetype was scored against one governing constraint, accountability is presumed non-subsumable, and sorted into four operating modes.
The four modes sit on a single axis: the more a seat is defined by routine workflow the further left it falls and the more subsumable it is; the more it is defined by accountability the further right, and the more it resists subsumption.
Beneath the matrix sits the finding that makes it work: across the sample, workflow intensity and accountability intensity are close to inversely correlated. The seats doing the most routine work carry the least institutional responsibility, and the seats carrying the most responsibility do the least routine work.
The archetype fates under that matrix, the roles subsumed to AI only, the archetypes restructured to AI plus human, and the class of human-only funds, which turned out to be empty, are the subject of the dispatch itself.
The Close
The reconstruction had begun with thirty-eight institutions.
It ended with a much smaller number of functions.
The next question was unavoidable
If institutions compress into functions, what happens when those functions meet artificial intelligence?
The full dispatch answers, function by function and archetype by archetype.
Back Matter
Research assets, intellectual property and access.
AI Subsumption Series · Back matter · AIS-05
Proprietary research assets
This publication is one visible output of a substantially larger proprietary research programme conducted by Odit Frontier Partners. The findings, frameworks and analysis presented here are anonymised by design; the underlying non-anonymised research corpus remains the exclusive intellectual property of Odit Frontier Partners.
That corpus includes, without limitation: non-anonymised organisational reconstructions of every firm in the sample; institutional architecture datasets; organogram reconstructions; canonical title mappings; decision-right mappings; governance mappings; reporting-line datasets; structural taxonomies; archetype classifications; institutional weight-class assignments; AI subsumability assessments; workflow and accountability classifications; confidence gradings; reconstruction metadata; evidence libraries; working papers; analytical notes; proprietary methodologies; and the prompts, reconstruction workflows and analytical tooling developed during the research programme. These assets constitute a proprietary institutional research corpus, and they are treated as such rather than as data incidental to a publication.
Access
The research assets described above are not available for sale. They are not publicly accessible, and they are not licensed separately from advisory work. Portions of the research corpus may be used only within carefully scoped advisory or research engagements, where doing so is necessary to deliver the agreed work, and only to that extent.
Engagement discretion
Making an enquiry or requesting an engagement does not create any entitlement to access these research assets. Odit Frontier Partners retains sole and absolute discretion to determine whether any engagement will be accepted, and likewise retains sole discretion to determine whether any portion of the underlying research corpus will be used within an accepted engagement. Odit Frontier Partners reserves the right to decline any engagement or request for access, in whole or in part, without being required to provide reasons or justification.
Reader interpretation
This publication represents the analysis of Odit Frontier Partners. Readers are free to agree or disagree with its conclusions, and the firm assumes no obligation to persuade every reader. The publication is offered for analytical consideration rather than consensus.
Disclosure
Odit Frontier Partners assumes no obligation to disclose proprietary datasets, reconstruction files, working papers, confidence gradings, internal methodologies, analytical workflows, prompts, intermediate analytical artefacts, or any other proprietary research assets beyond what it has chosen to publish. Readers are invited to evaluate the publication on its own merits; nothing in it creates any entitlement to further disclosure.
Back Matter
Methodology and claims.
This dispatch draws on a structural reconstruction of 38 investment firms across African markets, built from public sources: team pages, professional biographies, deal announcements, promotion histories, regulatory disclosures, appraisal documents published by development finance institutions, and the firms' own published materials. Reporting relationships that could not be confirmed are treated as inference throughout the underlying research and are graded accordingly. All organogram examples in Parts II and III are syntheses recombined from multiple firms; no example reproduces any single firm's structure, and the mapping from examples to source material is held privately. Firms are named only where they have publicly self-described the models discussed (Secha Capital, Kadan Capital, Ripple Ventures, Vela Partners, Moonfire Ventures, SignalFire), and all operational figures attributed to those firms, including agent counts, throughput multiples, cost comparisons and headcount-equivalence claims, are the firms' own public statements: they are self-reported, unaudited, and quoted here as claims rather than as findings. The full fund-by-fund reconstruction, the complete role-semantics ontology, the compression scoring and the institutional matrices are held as a research appendix.
About Odit Frontier Partners
Odit Frontier Partners (OFP) is a frontier capital architecture firm focused on the design of adaptive capital systems in volatile and emerging markets. The firm operates at the intersection of private capital, system design, and strategic foresight, building frameworks that enable capital to move, adapt, and compound under conditions of structural uncertainty.
About the author
Doris Odit Achenga is the founder of Odit Frontier Partners (OFP), a frontier capital architecture firm. Her work focuses on the design of adaptive capital systems in volatile markets.
Back Matter
Copyright, disclaimer and acknowledgements.
Copyright notice
© 2026 Odit Frontier Partners (OFP) Advisory Services SMC Ltd. All rights reserved.
This dispatch is the intellectual property of Odit Frontier Partners. No part of this work may be reproduced, distributed, transmitted, or stored in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without prior written permission, except for brief quotations used in reviews or academic reference. This work contains proprietary frameworks, concepts, and methodologies developed by Odit Frontier Partners, including but not limited to the minimum viable institutional structure, the weight-class taxonomy, the roles roster, the compression crosswalk and the four operating modes. Unauthorised use, replication, or commercial application of these materials is strictly prohibited.
Disclaimer
This dispatch is provided for informational and educational purposes only and does not constitute investment advice, legal advice, financial advice, or an offer to buy or sell any financial instrument. The views, frameworks, and strategies presented reflect the author's professional experience and analytical perspective at the time of writing. While every effort has been made to ensure conceptual integrity, no representation or warranty, express or implied, is made as to the completeness or reliability of the information contained herein. Readers are encouraged to exercise independent judgment and seek appropriate professional advice before making any investment or business decisions. Odit Frontier Partners (OFP) and the author shall not be held liable for any direct or indirect loss arising from the use or application of the concepts presented in this work. Certain frameworks and methodologies referenced in this dispatch are part of ongoing proprietary development and may not be fully disclosed.
Acknowledgements
This dispatch is part of the AI Subsumption series and applies the framework established in the series' opening dispatch, AI Subsumption and the New-Age Disruptor, to the fund industry itself. The reconstruction and its interpretation are the author's own.
Colophon
Publication details.
Odit Frontier Partners · Knowledge, IP & Systems Lab