AISOFT LLC Agentic Engineering

AISOFT Agentic Engineering brand system

Working brand system for the curriculum, workshop, and job-pathways material.

Direction: Stitch-style documentation. Calm, precise, useful, built around people doing real work.

Positioning

AISOFT helps teams ship useful AI work with agentic engineering.

Signal What it should communicate
Practical Learners leave with repos, commands, diffs, checks, and proof.
Practical judgment The material teaches review, gates, and delivery discipline.
In-person ready The workshop should feel like a room where people build, review, and improve together.

Brand idea

Clear work, visible proof.

The work spans agents, data, SDLC, MLOps, AI apps, workflows, and leadership adoption. The brand should make it feel clear, teachable, and executable.

Use language that sounds like:

  • “Here is the next step.”
  • “Here is the proof.”
  • “Here is what good looks like.”
  • “Here is how to recover when the agent drifts.”

Avoid language that sounds like:

  • “Unlock the future.”
  • “Revolutionize your workflow.”
  • “Magic AI productivity.”
  • “One guide to master everything.”

Visual System

Palette

Use the blue/slate system for the curriculum.

Token Value Use
Primary #004ac6 Links, active states, small emphasis
Primary strong #003996 Hover and primary action states
Primary soft #dbe8ff Very light emphasis backgrounds
Surface #f7f9fb Page background outside docs surfaces
Surface low #f1f5f9 Subtle panels and code backgrounds
Border #d9e2ef Dividers, tables, inputs, nav boundaries
Text #0f172a Primary text
Muted text #565e74 Supporting copy

Do not bring back the old warm curriculum palette. The logo can remain; the curriculum visual language is blue/slate.

Typography

Use Font Treatment
Headings Geist Strong, direct, no decorative tracking except small labels
Body Inter 16px baseline, generous line height
Code JetBrains Mono / ui-monospace Small, readable, copyable

Headings should be plain and literal.

Good:

Lesson 3.3 · Build it, step by step
Data + AI practice labs
Job pathways + AISOFT offerings

Avoid:

Unlocking the Agentic Future
Your AI Transformation Starts Here
The Complete AI Playbook

Layout

Use a documentation layout by default:

  • left navigation,
  • content column,
  • optional right-side table of contents,
  • minimal top navigation,
  • no decorative hero section on lesson pages.

The page should open on the material. Brand shows through restraint, consistency, and useful structure.

Components

Component Rule
Sidebar links Flat text links with a blue left active border. No pill cards.
Code blocks Border-led, light background, copy button, no heavy shadow.
Tables Simple borders, compact rows, no marketing-card treatment.
Callouts Use only when the learner needs a warning, decision, proof rule, or engineering judgment tip.
Buttons Use sparingly. Prefer links in docs; reserve buttons for actual actions.

Documentation Patterns

Engineering Judgment Tip

Use this when the learner needs judgment.

Engineering Judgment Tip

Optimize for readability over brevity. In agentic systems, the bottleneck is usually the human’s ability to review, debug, and recover from the agent’s choices.

Stitched Progress

Use a thin blue left rule to show the active concept, lesson, or step. Keep inactive steps plain.

State Treatment
Active Blue left rule, blue label, normal white background
Complete Muted text with proof attached
Later Muted text only; no decorative badge needed

Code Blocks

Every command should be copyable and readable without horizontal guessing.

git status --short
npm test

Voice

Voice: a capable guide working beside the learner.

Write like this:

Start with the smallest version. Run it on a fixture. Read the diff before you accept it.

Not like this:

This module empowers you to harness agentic workflows for unparalleled productivity.

Copy Rules

Use:

  • short sentences,
  • concrete nouns,
  • commands,
  • proof,
  • small next steps,
  • direct correction when needed.

Avoid:

  • hype,
  • vague transformation language,
  • dramatic framing,
  • meta explanations about the curriculum itself,
  • long intros before the learner reaches the task.

Workshop Feel

The in-person version should feel like a working room:

  1. brief setup,
  2. shared demo,
  3. learner runs the agent,
  4. diff review,
  5. proof command,
  6. next instruction.

Support that cadence. Make it easy to ask:

  • What changed?
  • How did you verify it?
  • What would you ask the agent next?

Reusable Copy

One-line Description

A practical agentic engineering curriculum for teams that need to brief, build, review, evaluate, and ship AI-assisted work.

Workshop Description

Agentic Engineering Day is an in-person workshop where learners use agents on real projects, review real diffs, run proof commands, and leave with a first-week adoption plan.

Short CTA

Bring a repo. Build one slice. Leave with proof.

Quality Bar

Do not call it done until the diff is read, the check has passed, and the proof is attached.

Brand Checklist

Before publishing a new page, check:

  1. Does the page start with the useful material?
  2. Is the first task or decision visible quickly?
  3. Are commands shown in copyable blocks?
  4. Is the tone calm and specific?
  5. Are there any marketing phrases that can be replaced with proof?
  6. Does the page use blue/slate docs styling, not the old warm palette?
  7. Would a fresh graduate and an experienced engineer both know what to do next?