SmartArt-style diagrams for Typst posters, papers and presentations: flowcharts, process, cycle, hierarchy, relationship, matrix and pyramid diagrams, plus data widgets (bars, gauges, waffle, KPI cards), timelines, Gantt charts and tables.

• Fast. Every widget is drawn with native Typst primitives and has no dependencies; a diagram costs a few milliseconds, and a flowchart is about 40× faster than the same diagram in fletcher.
• Text always fits. Boxes grow to fit their longest label; round and slanted shapes keep their size and shrink all labels to one shared size that fits, as SmartArt does.
• Connectable components. flowchart composes nodes, chained edges and titled groups, laid out on a grid or automatically; with panels and loop-ring it rebuilds typical paper figures (see Paper figures).

Installation

#import "@preview/typart:0.2.1": *

Items and colours

Unless noted otherwise, every widget takes a list of items where each item is either a label (content) or a (label, color) tuple that overrides that item’s colour. Items without an explicit colour cycle through the built-in palette (lavender, green, amber, red, sky), which is also exported for your own use.

Dimension parameters given as plain numbers (w, h, radius, …) are centimetres; they also accept any Typst length (2.5cm, 40pt, …). Parameters documented with pt/cm defaults are regular Typst lengths.

Sizes documented as heights are minimums: boxes grow to fit their text. Spacings that carry arrows (stairs, arrows, cycle, radial) never shrink below what keeps every arrow visible.

Flowcharts and pipelines

flowchart builds method, pipeline and decision diagrams from three components:

• node(name, [(column, row)], body, ..style) — a box, pill, circle, ellipse, diamond, loop or no shape at all (shape: none) with any content (icons, images, math), and an optional caption hanging below it. shape: none gives a bare icon or plot.
• edge("a", "b", "c", ..) — connects nodes (or named groups) in a chain: a → b → c. Routing is automatic: straight when the two shapes face each other, elbowed otherwise, and around the row when the straight line is blocked or would overlap a reverse edge. via: forces a route through grid points, bend: curves the edge.
• group(title, "a", "b", ..) — a dashed frame around nodes, titled on its border or inside: true; give it a name to connect edges to the whole frame.

Nodes with a (column, row) position sit on a grid whose columns and rows take the size of their largest node; gaps widen to fit group frames, and leave room for an arrow wherever an edge crosses them; fractional positions such as (0.5, 1) sit between tracks (centred over two nodes) and widen them when needed. Leave the position out and the layout is automatic: each node goes one layer after its longest chain of predecessors (topological order), and is aligned with its predecessors inside its layer.

See Paper figures for complete diagrams built with these pieces.

#flowchart(
  node("data", [Raw data]), node("clean", [Cleaning]), node("feat", [Features]),
  node("a", [Model A]), node("b", [Model B]), node("ens", [Ensemble]),
  node("out", [Report], shape: "pill", fill: rgb("#a7dd9b")),
  edge("data", "clean", "feat", "a", "ens", "out"),
  edge("feat", "b", "ens"),
  group([Training], "a", "b", "ens"),
)

A decision flowchart, top to bottom, with edge labels and a dashed back edge:

#flowchart(
  dir: ttb,
  node("start", [Start], shape: "pill", fill: rgb("#a7dd9b")),
  node("read", [Read input]),
  node("ok", [Valid?], shape: "diamond", fill: rgb("#f9c74f")),
  node("run", [Process]),
  node("err", [Report error], fill: rgb("#f38ba8")),
  node("end", [End], shape: "pill", fill: rgb("#a7dd9b")),
  edge("start", "read", "ok"),
  edge("ok", "run", label: [yes]),
  edge("ok", "err", label: [no]),
  edge("run", "end"),
  edge("err", "end"),
  edge("err", "read", dash: "dashed", label: [retry]),
)

A research pipeline placed on the grid, with groups (nodes may hold any content, e.g. image(...) plots):

#set text(size: 9pt)
#let (key, model, feature, feat, eval) = (rgb("#8f8fde"), rgb("#a9a9e6"), rgb("#f5b57a"), rgb("#89b4fa"), rgb("#f38ba8"))
#flowchart(
  spacing: (12mm, 12mm),
  node("pwd", (3, -1), [User password], fill: key),
  node("sha", (4, -1), [SHA-256], fill: key),
  node("map", (5, -1), [Chaotic map \ (logistic / tent)], fill: model),
  node("smat", (6, -1), [Projection \ $S in RR^(d times n)$], fill: model),
  node("ecg", (0, 0), [ECG signal], fill: white),
  node("ppg", (0, 1), [PPG signal], fill: white),
  node("dct1", (1, 0), [DCT], fill: feature),
  node("dct2", (1, 1), [DCT], fill: feature),
  node("cat", (2, 0), [Concatenate \ first halves], fill: white),
  node("idct", (3, 0), [IDCT], fill: feature),
  node("mfcc", (4, 0), [MFCC], fill: feat),
  node("x", (5, 0), [Features \ $x in RR^n$], fill: feat),
  node("tpl", (6, 0), [Template \ $t = S x$], fill: model),
  node("match", (6, 1), [Cosine matcher], fill: eval),
  node("metrics", (5, 1), [AUC · EER · DET], fill: white),

  edge("pwd", "sha", "map", "smat", "tpl"),
  edge("ecg", "dct1", "cat", "idct", "mfcc", "x", "tpl", "match", "metrics"),
  edge("ppg", "dct2", "cat"),

  group([Chaotic projection], "pwd", "sha", "map", "smat"),
  group([Signal fusion], "dct1", "dct2", "cat", "idct", title-pos: bottom),
  group([Feature extraction], "mfcc", "x"),
  group([Evaluation], "match", "metrics", title-pos: bottom),
)

flowchart

node

edge

group

Paper figures

Building blocks for figures in papers and posters: panels lays out framed sub-figures with captions and a), b)… labels; loop-ring draws a “repeat / continuous” ring around any content; flowchart composes method diagrams. The figures below are complete examples, set in Inter (any installed font works); icons are Font Awesome glyphs.

Images. A node body, a panels item and a loop-ring accept any content, so a picture is just image(...); the path is resolved from your own document. Combine it with shape: none for a bare picture, and with caption: for a label under it:

#flowchart(
  node("sig", (0, 0), image("images/ecg.svg", width: 3cm), shape: none, caption: [ECG]),
  node("model", (1, 0), [Model]),
  edge("sig", "model"),
)
#panels(columns: 2, (image("images/ecg.svg", width: 4cm), [Raw signal]), ([Filtered], [Caption]))

A method diagram in three lanes: independent-columns: true lets each lane stack its own rows, captions hang under the boxes, fractional positions centre a node over two columns, routes go through grid points (via), arrows reaching the same side of a node get their own slots, stretch grows the K-Fold frame to the right and the Validation frame down until they meet, and join draws both as one outline:

#set text(font: "Inter", size: 16pt, weight: "bold")
#let (green, red, blue, orange, violet) = (rgb("#a7dd9b"), rgb("#f38ba8"), rgb("#89b4fa"), rgb("#f9b98a"), rgb("#b4bcfa"))

#flowchart(
  node-stroke: 1.5pt + black,
  title-style: (weight: "regular", fill: luma(110), size: 18pt),
  caption-style: (size: 10pt),
  independent-columns: true,
  spacing: (12mm, 9mm),
  edge-stroke: 1.5pt + black,
  radius: 10pt,
  inset: 10pt,
  node("dataset", (0.5, 0), [Dataset], fill: green, caption: [Dataset Name \ Users: Number]),
  node("load", (0.5, 1), [Load Data], fill: green, caption: [Num. Segments: N \ Size segment: Fs\*t]),
  node("sig1", (0, 2), [Signal1], fill: red),
  node("sig2", (1, 2), [Signal2], fill: blue),
  node("seg", (0.5, 3), [Segment-by-segment \ preprocessing], fill: green),
  node("pre1", (0, 4), [Signal1 \ Prepo], fill: red.darken(10%)),
  node("pre2", (1, 4), [Signal2 \ Prepo], fill: blue.darken(10%)),

  node("feat", (2.5, 0), [Feature extraction], fill: orange,
    caption: [Wavelet: bior3.3 \ characteristics by coefficient: \ [mean, std, var, kurtosis, skewness, \ energy, entropy, max, min, median, iqr, range, mad]]),
  node("f1", (2, 1), [Signal1 F], fill: red.lighten(50%)),
  node("f2", (3, 1), [Signal2 F], fill: blue.lighten(50%)),
  node("merge", (2.5, 2), [Merging features], fill: orange),
  node("final", (2.5, 3), [Final features], fill: rgb("#f5c2e7")),
  node("kfold", (2.5, 4), [Stratified K-Fold], fill: violet, caption: [Folds: 5 \ [Fold 1] [Fold 2] … [Fold k]]),

  node("list", (4, 0), [List of classifiers], fill: violet,
    caption: [classifiers: [Random Forest, Logistic \ Regression, KNN, Decision Tree, \ Naive Bayes, Extra Trees, LDA]]),
  node("cv", (4, 1), [Cross-Validation loop], fill: violet, caption: [Train and test classifier by classifier]),
  node("report", (4, 2), [Classification Report], fill: violet, caption: [Save in parquet packages]),
  node("viz", (4, 3), [Visualization], fill: violet),

  edge("dataset", "load"),
  edge("load", "sig1", "seg", "pre1"),
  edge("load", "sig2", "seg", "pre2"),
  edge("pre1", "feat", via: ((0, 4.7), (1.6, 4.7), (1.6, 0))),
  edge("pre2", "feat", via: ((1, 4.55), (1.5, 4.55), (1.5, 0))),
  edge("feat", "f1", "merge", "final", "kfold"),
  edge("feat", "f2", "merge"),
  edge("kfold", "list", via: ((3.45, 4), (3.45, 0))),
  edge("list", "cv", "report", "viz"),

  group([Preprocessing], "dataset", "load", "sig1", "sig2", "seg", "pre1", "pre2", inside: true),
  group([Feature Engineering], "feat", "f1", "f2", "merge", "final", inside: true),
  group(none, "kfold", name: "kfold-box", stretch: right),
  group([Validation and Results], "list", "cv", "report", "viz", inside: true, stretch: bottom,
    join: "kfold-box"),
)

A continuous-authentication scheme: icons as nodes, edges between whole groups, a decision diamond, a loop node and a curved back edge:

#set text(font: "Inter", size: 12pt, weight: "bold")

#flowchart(
  node-stroke: 1.2pt + black,
  title-style: (weight: "regular", fill: luma(110), size: 14pt),
  spacing: (14mm, 16mm),
  edge-stroke: 2.5pt + black,
  radius: 8pt,
  node("user", (0, 0), icon("\u{f007}", fill: rgb("#64b5f6"), size: 36pt), caption: [Sesion \ Request], shape: none),
  node("auth", (1, 0), [Authentication \ Request], fill: rgb("#a7dd9b")),
  node("watch", (2, 0), icon("\u{f21e}", fill: rgb("#f44336"), size: 44pt), shape: none),
  node("capture", (2, 1), [Signal Capture], fill: rgb("#f38ba8")),
  node("plot", (3, 1), icon("\u{f201}", fill: rgb("#e53935"), size: 44pt), shape: none),
  node("convert", (4, 0), [Convert to \ Features], fill: rgb("#f9b98a")),
  node("vector", (4, 1), text(weight: "regular", size: 16pt, $[F_1, F_2, ..., F_n]$), shape: none),
  node("model", (4, 2), [Pre-trained \ model], fill: rgb("#b4bcfa")),
  node("classify", (4, 3), [Classify], fill: rgb("#b4bcfa")),
  node("check", (3, 2), [Is #icon("\u{f007}", fill: rgb("#64b5f6"), size: 22pt) ?], shape: "diamond", fill: rgb("#f6dfac"), stroke: 1pt + black),
  node("again", (2, 2), [New session \ time], shape: "loop", fill: white, stroke: 5pt + rgb("#a7dd9b")),
  node("ended", (3, 3), icon("\u{f506}", fill: rgb("#f06292"), size: 36pt), caption: [Session ended], shape: none),

  edge("user", "auth", "capture-box"),
  edge("watch", "capture", "plot"),
  edge("capture-box", "features"),
  edge("convert", "vector"),
  edge("features", "model-box"),
  edge("model", "classify"),
  edge("model", "check"),
  edge("check", "again", label: text(weight: "regular")[yes]),
  edge("check", "ended", label: text(weight: "regular")[no]),
  edge("again", "auth", bend: 45deg),

  group([Continuous signal capture], "watch", "capture", "plot", name: "capture-box", inside: true),
  group([Feature extraction], "convert", "vector", name: "features", inside: true),
  group([Model], "model", "classify", name: "model-box", inside: true),
  group([Verification module], "again", "check", inside: true),
)

A verification pipeline with an image node (images/ecg.svg), edges between whole groups and two branches leaving one node through waypoints:

#set text(font: "Inter", size: 13pt, weight: "bold")

#flowchart(
  node-stroke: 1.2pt + black,
  title-style: (weight: "regular", size: 15pt),
  spacing: (10mm, 9mm),
  edge-stroke: 2.5pt + black,
  radius: 8pt,
  group-stroke: (paint: luma(120), thickness: 1pt, dash: "dashed"),
  node("user", (0, 0), icon("\u{f007}", fill: rgb("#c48b9f"), size: 44pt), caption: [Sesion \ Request], shape: none),
  node("watch", (0, 1), icon("\u{f21e}", fill: rgb("#f44336"), size: 52pt), shape: none),
  node("capture", (0, 2), [Capture], fill: rgb("#f38ba8"), width: 4cm),
  node("ecg", (1, 2), image("images/ecg.svg", width: 3cm), shape: none),
  node("feat", (2, 1), [Feature extraction], fill: rgb("#f9b98a"), width: 4.6cm, height: 1.6cm),
  node("prep", (2, 2), [Preprocesing], fill: rgb("#f9b98a"), width: 4.6cm, height: 1.6cm),
  node("rejected", (1, -1), icon("\u{f235}", fill: rgb("#f06292"), size: 44pt), caption: [Rejected], shape: none),
  node("or", (2, -1), text(weight: "regular", size: 18pt)[OR], shape: none),
  node("accepted", (3, -1), icon("\u{f4fc}", fill: rgb("#c48b9f"), size: 44pt), caption: [Accepted], shape: none),
  node("auth", (2, 0), [Autentication], fill: rgb("#b4bcfa"), width: 4.6cm),

  edge("user", "capture-box", "features", "verification"),
  edge("watch", "capture", "ecg"),
  edge("prep", "feat"),
  edge("auth", "rejected", via: ((1.75, -1),)),
  edge("auth", "accepted", via: ((2.25, -1),)),

  group([Signal Capture], "watch", "capture", "ecg", name: "capture-box", inside: true),
  group([Feature Engineering], "feat", "prep", name: "features", inside: true),
  group([Verification Stage], "rejected", "or", "accepted", "auth", name: "verification", inside: true),
)

Sub-figure grids with panels:

#set text(font: "Inter", size: 12pt)

#panels(columns: 3, numbering: "1)",
  (hstack(icon("\u{f023}", fill: rgb("#8bc34a")), icon("\u{f27a}", fill: rgb("#2196f3")), icon("\u{f06e}", fill: rgb("#26a69a"))), [Something a user knows]),
  (hstack(icon("\u{f3cf}", fill: rgb("#29b6f6")), icon("\u{f084}", fill: rgb("#ffc107")), icon("\u{f2c2}", fill: rgb("#2196f3"))), [Something in a user's possession]),
  (hstack(icon("\u{f577}", fill: rgb("#29b6f6")), icon("\u{f007}", fill: rgb("#90a4ae")), icon("\u{f06e}", fill: rgb("#00897b"))), [Something the user is]),
  (hstack(icon("\u{f554}", fill: rgb("#e53935")), icon("\u{f11c}", fill: rgb("#e53935")), icon("\u{f25a}", fill: rgb("#1e88e5"))), [Something the user does]),
  (hstack(icon("\u{f3c5}", fill: rgb("#ef5350")), icon("\u{f1eb}", fill: rgb("#1e88e5"))), [Somewhere that a user is]),
  (hstack(icon("\u{f519}", fill: rgb("#1e88e5")), icon("\u{f1bb}", fill: rgb("#43a047"))), [Something that is in the user's environment]),
)

#set text(font: "Inter", size: 12pt)
#let op(t) = text(size: 22pt, t)

#panels(columns: 2,
  (hstack(icon("\u{f023}", fill: rgb("#8bc34a")), op[OR], icon("\u{f577}", fill: rgb("#29b6f6")), op[OR], icon("\u{f084}", fill: rgb("#ffc107"))), [Single-Factor]),
  (hstack(icon("\u{f06e}", fill: rgb("#26a69a")), op[And], icon("\u{f577}", fill: rgb("#29b6f6")), op[OR], icon("\u{f3c5}", fill: rgb("#ef5350"))), [Multi-Factor]),
  (hstack(icon("\u{f83e}", fill: rgb("#e53935")), op[Merge], icon("\u{f577}", fill: rgb("#29b6f6"))), [Multimodal]),
  (loop-ring(hstack(icon("\u{f21e}", fill: rgb("#e53935")), [Verification]), stroke: 4pt + rgb("#a7dd9b")), [Continuous]),
)

panels

icon and hstack

Two helpers for figure content: icon sets a glyph from an icon font (Font Awesome 7 Free by default, used by the examples above), and hstack puts items side by side on a common middle line.

#hstack(icon("\u{f023}", fill: rgb("#8bc34a")), text(size: 22pt)[OR], icon("\u{f084}", fill: rgb("#ffc107")))

loop-ring

List

steps

A vertical numbered list of steps.

#steps(([Plan], [Execute], [Review], ([Deploy], rgb("#e64553"))))

hlist

A horizontal row of equally sized coloured blocks.

#hlist(([Plan], [Execute], ([Review], rgb("#e64553"))))

cards

Columns with a coloured header over a tinted body (horizontal bullet list). Headers share one height, and so do bodies.

#cards((
  ([Goals], list([Faster delivery], [Lower cost])),
  ([Risks], list([Scope creep], [Budget overrun]), rgb("#e64553")),
  ([Actions], [Weekly review of progress against the plan.]),
))

Items are tuples (title, body) or (title, body, color).

swot

A SWOT analysis as a 2×2 of cards (built on cards).

#swot(
  [- Experienced team \ - Unique dataset],
  [- Small budget],
  [- New market \ - Open data],
  [- Competition],
)

hexagons

A zig-zag strip of nested hexagons (honeycomb).

#hexagons(([Idea], [Design], ([Build], rgb("#e64553")), [Launch], [Learn]))

Process

flow

Blocks joined by block arrows (basic process), horizontal or vertical.

#flow(([Input], [Process], [Review], ([Output], rgb("#e64553"))))
#flow(([Input], [Process], [Output]), dir: ttb)

chevron

A sequence of right-pointing chevron arrows.

#chevron(([Phase 1], [Phase 2], [Phase 3], [Phase 4]))

snake

A bending process that wraps into multiple rows.

#snake(([Step 1], [Step 2], [Step 3], [Step 4], [Step 5], [Step 6]))

stairs

An ascending stair-step process.

#stairs(([Plan], [Develop], ([Launch], rgb("#e64553"))))

timeline

A horizontal timeline with alternating labels above and below the line.

#timeline((("2024", [Start]), ("2025", [Pilot]), ("2026", [Launch])))

Items are tuples (date, label) or (date, label, color).

vtimeline

A vertical timeline with cards alternating left and right of a central spine.

#vtimeline((
  ("2023", [Pilot study with 20 participants]),
  ("2024", [Data collection across three hospitals], rgb("#e64553")),
  ("2025", [Journal paper]),
  ("2026", [Product launch]),
))

Items are tuples (date, body) or (date, body, color).

stepper

Numbered milestones on a line, marking what is done, what is in progress and what is next.

#stepper(([Proposal], [Data collection], [Model], [Paper], [Defense]), current: 2)

gantt

A Gantt chart with task bars, progress indicator, grid lines, and a “today” marker.

#gantt(
  (
    ("Planning",     0, 12, 100),
    ("Development",  6, 30, 60),
    ("Testing",     24, 40, 30),
    ("Papers",       0, 48, 20, false),
  ),
  span: 48, grid-step: 12,
  periods: ("Year 1", "Year 2", "Year 3", "Year 4"),
  today: 20,
)

Each task is a tuple (name, start, end), (name, start, end, progress) or (name, start, end, progress, group) — group: false renders the bar greyed instead of in the accent colour.

Cycle

cycle

Nodes arranged on a ring connected by curved arrows.

#cycle(([Plan], [Do], [Check], ([Act], rgb("#e64553"))))

donut

A ring split into labelled segments (segmented cycle), clockwise from the top.

#donut(([Plan], [Build], [Test], ([Ship], rgb("#e64553"))), center: [Cycle])

Hierarchy

hierarchy

A root node with a row of children beneath it.

#hierarchy([CEO], (([Eng], rgb("#a7dd9b")), [Sales], [HR]))

tree

A multi-level tree hierarchy. A node is a label, (label, children), (label, color), or ((label, color), children). Box heights grow automatically to fit wrapped text.

#tree(([CEO], (([Eng], ([Web], [Mobile])), [Sales])))

Relationship

venn

A 2- or 3-circle Venn diagram.

#venn((([Behavior], rgb("#aab4f7")), [Signal], [Security]))

target

Concentric rings (bullseye), outermost item first.

#target(([Vision], [Mission], [Goal]))

radial

A centre circle joined by spokes to satellite circles.

#radial([Goal], ([People], [Process], ([Tools], rgb("#e64553")), [Data]))

arrows

Converging or diverging arrows around a central core.

// converging
#arrows(([Idea 1], [Idea 2], [Idea 3]), core: [Core])

// diverging
#arrows(([Branch 1], [Branch 2]), diverge: true)

opposing

Two block arrows facing each other.

#opposing(([Pros], [Cons]))

equation

An A + B = C equation-style diagram (last item is the result).

#equation(([Input 1], [Input 2], [Output]))

gears

Up to 3 interlocking cogwheels (extra items are ignored).

#gears(([Research], [Development], [Production]))

Matrix and pyramid

matrix

A 2x2 matrix with optional axis labels. Cells in order: top-left, top-right, bottom-left, bottom-right.

#matrix(([TL], [TR], [BL], [BR]),
  x-axis: ("Low", "High"),
  y-axis: ("High", "Low"))

pyramid

A pyramid or funnel diagram (use flip: true for a funnel).

#pyramid(([Vision], [Strategy], [Tactics]))

Data and poster

kpi

Stat cards displaying a large value and a caption.

#kpi((("< 1%", [Target EER]),
      ("99%", [Accuracy], rgb("#40a02b")),
      ("500+", [Users])))

Items are tuples (value, caption) or (value, caption, color).

bars

Labelled horizontal bars with their values: results tables made visual. A value that does not fit inside its bar is drawn right after it.

#bars(
  (([Ours], 94.2, rgb("#e64553")), ([Baseline A], 88.7), ([Baseline B], 81.3), ([Baseline C], 12.5)),
  unit: "%", max: 100,
)

Items are tuples (label, value) or (label, value, color).

gauges

Progress rings with a big percentage and a caption.

#gauges(((72, [Recall]), (91, [Precision], rgb("#e64553")), (100, [Coverage])))

Items are tuples (percent, caption) or (percent, caption, color).

waffle

A grid of squares showing shares of a whole, with a legend; cells beyond the parts stay grey.

#waffle(([Responders], 62), ([Partial], 23), ([No change], 15, rgb("#e64553")))

Each part is (label, count) or (label, count, color).

pill-steps

A central hub with connected pill-shaped step rows. Requires Font Awesome 7 Free for icons.

#pill-steps(
  (
    ([Research], [Gather requirements],
      icon("\u{f0eb}", size: 26pt)),
    ([Develop],  [Build the solution],
      icon("\u{f0ae}", size: 26pt), rgb("#74c7ec")),
  ),
  title: [PROJECT STEPS],
)

Items are tuples (title, body), (title, body, icon), or (title, body, icon, color). The icon may be a Font Awesome glyph string or any pre-styled content. Pill heights grow automatically to fit wrapped text.

process

A heartbeat-ring process diagram with icon badges and connector arms. Requires Font Awesome 7 Free.

#process(
  (
    ("\u{f02d}", "1. Review methods"),
    ("\u{f51e}", "2. Data collection"),
    ("\u{f83e}", "3. Signal analysis"),
  ),
)

Items are up to 6 tuples (icon-glyph, label) or (icon-glyph, label, color); the first three render on the left, the last three mirrored on the right. Any Font Awesome codepoint works as the glyph.

card-table

A card-style table with rounded cells, coloured header, and alternating row fills.

#card-table(
  ("Study", "Approach", "Dataset"),
  (
    ("Guo et al.",  "Propose two...", "Homemade"),
    ("Gurunathan",  "Exhaustive...", "PUT Vein"),
    ("Zhang et al.", "ML-based",     "InHouse"),
  ),
)

To override a single row’s colours, wrap it with card-row:

#card-row(("This work", "Ours", "New"), fill: rgb("#a7dd9b"), text-fill: white)