The visual and verbal identity for the African Center for Data Science and Analytics — a research, consulting, and training institute bridging the continental skills gap in data, AI, and analytics.
Positioning · Audience · Promise
AfriCDSA sits between the academy and the boardroom. The identity has to carry the weight of both: the authority of a certification body and the clarity of a modern consultancy.
To close the African data-skills gap through accredited training, applied research, and consulting rooted in local context.
Early-career professionals seeking certification; mid-career managers upskilling into analytics; ministries, NGOs and enterprises buying consulting.
Credentials recognized globally. Methods grounded locally. Decisions that hold up under scrutiny.
Tagline — lockup
Always set lowercase. Italic on "to decisions". Comma is required — the pause matters.
Primary mark · Variants · Clear space · Misuse
The primary lockup pairs an orange continent-silhouette (overlaid with a data-network motif) with the "AfriCDSA" wordmark and the "from data to decisions" tagline. Use as a complete lockup whenever possible.
A contemporary alternative designed from the ground up to scale for digital surfaces, favicons, and small applications. The monogram reads as an "A" built from data nodes — directly encoding the institute's discipline into the mark. Pairs with a custom wordmark set in Newsreader.
The monogram replaces the continent-silhouette with a structural reading: six nodes arranged as the letter "A", connected by edges. The apex node is flamingo — the "decision" — with a central sand node representing the data foundation.
It scales cleanly from 16px (favicon) to 4m (signage), reads in monochrome, and avoids the cliché of repeating a map-of-Africa in every mark. Use v2 in digital contexts, product UI, and small-scale applications. Use v1 for formal certificates and legacy contexts where the existing mark is already recognized.
Maintain a minimum clear space of X — the cap-height of the "A" — on all sides. No element, image, or edge may enter this zone.
— don't distort proportions
— don't place on clashing gradients
— don't recolor the mark
Core palette · Scales · Semantic
Three colors do the heavy lifting. Navy is dominant — it carries surfaces, headers, and typography. Flamingo is reserved for moments of decision: calls to action, highlights, a single italic word. Sand warms the system and softens academic density.
Navy-600 on white passes AAA (contrast 14.2:1). Flamingo-500 on white meets AA-large but fails AA for small body text — restrict it to headlines, accents and 18pt+ copy. For interactive elements on white, use Flamingo-700 (#AE350B). Sand-400 is never used for text on white — reserve it for surfaces.
Display · Body · Mono
Newsreader carries editorial authority — its italic variant is the voice of emphasis. IBM Plex Sans handles every functional surface: navigation, body, forms, dashboards. Plex Mono is reserved for data, code, and metadata labels.
The quick brown fox jumps over the lazy dog. 1234567890. Newsreader is used at 32px and above; below that it loses its authority. Italic variants are used sparingly, as an emphatic voice within a running line.
Plex Sans is the primary workhorse. Use weight 400 for body, 500 for interface labels, 600 for emphasis. Letter-spacing stays at defaults; never condense. Plex's OpenType SS01/SS02 features are enabled system-wide for the rounded single-storey "a" and "g" — this softens the voice.
Plex Mono is reserved for code, data tables, metadata labels, eyebrows, and small all-caps captions. Always set with letter-spacing 0.08–0.14em for labels, 0 for code.
| Display / 92-72px / 0.98 / -0.025em | From data, to decisions. |
| H1 / 44px / 1.05 / -0.02em | Evidence-led methods. |
| H2 / 32px / 1.15 / -0.015em | Certification pathways |
| Lede / 22px / 1.45 / 300 wt | A lede paragraph sets the thesis of a section. |
| Body / 15px / 1.55 / Plex 400 | Body copy uses IBM Plex Sans at 15 over 1.55. Measure capped at 68 characters. |
| Small / 13px / 1.5 / Plex 400 | Captions, footnotes, and form helper text. |
| Eyebrow / 11px / 0.14em / Mono | Section · Label · Metadata |
Pattern system · Data-mesh · Graphic language
Every graphic device derives from one idea: information visualized. Nodes, connections, grids, iso-curves, and gradients — each one is a shorthand for a method the institute actually teaches. Used together, they form a recognizable signature.
The banner combines three devices — mesh, flow-lines, node-accents — against a navy field. This is the canonical pairing for editorial covers, report spreads, and hero sections.
The Token / Stream device visualises the atomic unit of language AI: the token. Variable-width bricks represent tokens of differing lengths — short function words, long technical terms — flowing in rows like a real token sequence. Colour encodes token category: flamingo marks high-signal content tokens, navy carries structural or stop tokens, sand flags numeric or named entities, and iris marks sub-word pieces. Use this device wherever the brand touches NLP, LLM, or text-analytics contexts.
Base layer ~55%. Structural tokens.
High activity ~20%. Content tokens.
Medium ~10%. Rare words, entities.
Secondary ~8%. Boundary tokens.
Accent ~7%. Sub-word pieces.
Scale the brick height with --token-bh (default 13px) and gap with --token-vg (default 4px). On dark fields only — never render on light or sand backgrounds.
Default spacing 22px. Tighten to 14px for data surfaces, loosen to 32px for covers and print.
Patterns are always single-hue on a single-hue field. Never more than two tones per pattern. Flamingo only appears as a point accent, never as a field.
Patterns sit behind content, faded 35–65%. Content is never placed directly over high-contrast nodes — use a mask or radial fade.
Adjust density with the --pattern-density token. The Tweaks panel exposes three density presets for live preview.
Style · Grid · Weight
Icons are drawn on a 24×24 grid at a uniform 1.5pt stroke, flat caps, mitered corners. Only the two brand accents appear in color — everything else is rendered in Navy or Ink. No fills, no gradients, no duotones.
Library includes pictograms for each certification (CDS, CDA, AI Expert, BA-Managers, DPO) and functional icons for the LMS. Export as SVG with stroke-as-path for scale fidelity.
Photography direction · Do & don't
Imagery is documentary — shot on location in actual classrooms, labs, boardrooms. Natural light preferred; never composited. People are shown working, not posing. A subtle navy duotone can be applied for editorial contexts, but only to single-subject portraits; leave group shots and contextual scenes in full color.
Chart palette · Conventions · Example charts
A data-science institute must exemplify the craft. Charts use restrained palettes — never the full rainbow. Categorical data uses the 6-step brand series. Sequential and divergent palettes are derived from Navy and Flamingo. Gridlines are dashed at 20% opacity. Axes are the faintest rule. Labels are set in Plex Mono at 9pt.
Sequential · Navy 50 → 800. Use for intensity, counts, heatmaps.
Divergent · Flamingo ↔ Navy. Use for +/- deltas, sentiment, bipolar scales.
Personality · Examples · Words we avoid
We write the way a senior researcher briefs a minister: clearly, with numbers when they help, never oversold. The brand is confident without being corporate, warm without being casual.
avoid: world-class · cutting-edge · leverage · synergize · revolutionary · disruptive · AI-powered (without specifics) · unlock · unleash · ecosystem (when it means "group") · game-changing
prefer: accredited · applied · certified · evidence · methods · outcomes · cohort · case · region-specific · measured
Stationery · Social · Digital
This is to certify that the named participant has successfully completed the Certified Data Scientist (CDS) programme, delivered by the African Center for Data Science and Analytics and accredited by the International Association for Business Analytics Certification (IABAC).
The programme comprised 240 hours of instruction, three capstone projects, and a terminal evaluation conducted under observed conditions.
The scope of the proposed engagement covers three workstreams: (1) diagnostic of current data maturity; (2) capacity-building for the analytics function; (3) implementation of a decision-support dashboard.
A phased 14-week plan is attached as Annex A.