Semanti design system.
A brand and interface system for an AI advisory, labs, and assurance firm built for regulated, high-stakes environments. Every element passes a ten-second credibility test. No stock imagery. No consultancy-speak. No urgency tricks.
Contents.
Navigate through the sidebar. Each section is granular and intended as the canonical reference.
01 · Brand brain
Core, statement, principles, users02 · Brand
Logos, marks, lockups
03 · Color
Palette, ramps, pairings, gradient04 · Type
Satoshi + Work Sans hierarchy05 · Spacing & form
8pt grid, radii, borders06 · Components
Button, link, nav07 · Voice
Do / don't, canon, mechanics08 · Agent
Brand as a system promptWhat Semanti stands for.
Brand alignment across three dimensions: goals, users, identity. Brand Core distills the most essential building blocks into a strategic foundation for identity and communications. Six categories, one attribute each. Everything that follows is downstream.
Brand core.
Six attributes. Each answers one question about how Semanti shows up in the world.
How would your community describe you?
Trusted-expertsHow would you describe your customers?
LeveragedHow do we sound to others?
IntelligentHow do users feel after interacting with us?
ReassuredWhat tangible impact do you have on others?
ClarityWhat makes you radically different?
LineageBrand statement.
The canonical paragraph. Frozen copy. Use verbatim in investor decks and engagement letters.
In regulated, high-stakes environments, AI decisions are not theoretical. They carry financial, regulatory, and reputational consequences. In this context, capability is not enough. Judgement is required — the kind that cannot be compressed, acquired, or simulated. Only built through experience.
Semanti gives leveraged leaders and operators clarity when AI decisions carry real risk. Our advisory, prototyping, and assurance are grounded in lived experience — direct exposure to the full arc of modern AI, from early adoption to scaled, governed deployment.
They come to us for direction and leave with conviction shaped by that experience. Something no new entrant can replicate.
Communication principles.
Six rules for how Semanti sounds, feels, and shows up. Every message, deck, and conversation filters through these to stay consistent, credible, and unmistakably Semanti.
Show, don't claim.
Operate from earned authority. Every piece of communication should be traceable to a real engagement, deployment, or perspective earned in the field. Trust is built through what's been done. That lineage is the only credential that matters.
Respect what's at stake.
The audience operates under real constraint — investor pressure, regulatory scrutiny, time compression. Never assume a lack of knowledge or trivialize their position. Speak to the consequences they navigate, not a gap in understanding.
Think out loud, with precision.
Sound like someone three steps ahead — who knows exactly what matters now. No filler, no consultancy-speak, no jargon for its own sake. Every sentence earns its place or gets cut.
Calm the room, then move it.
After every interaction, the client should feel like someone finally understood their problem and showed them a way through it. Don't add to the noise. Don't add to the panic. Reassurance isn't softness — it's proving there's solid ground beneath their feet.
Strip it until only the decision remains.
Every deliverable, conversation, and piece of content should leave the client with fewer questions and more conviction. If they still need to interpret what we gave them, we haven't finished. Clarity isn't simplification — it's elimination of everything that doesn't move them forward.
"We were there" always beats "we've studied it."
Speak from the timeline — raw prompts through to governed production deployment. Reference the real arc. This experience can't be compressed, acquired, or simulated. Use it as proof, not posture.
The compression window.
Why timing is the strategy.
Cloud 2015 → AI 2023. Same arc. Compressed timeline.
In 2015, HSBC asked Amazon why a bookshop was selling them servers. Within five years, they were running business-critical applications in the cloud and exiting data centers. The AI adoption curve follows the same stages — but compressed. What took cloud five years will take AI three.
The firms that won in cloud picked a proposition, built it, earned their scars, and scaled from there. They didn't start with a strategy deck. They started with proof. Semanti's lineage runs through both curves. That experience can't be compressed.
AI advisory for private equity will compress fast. The winners won't be those with the best methodology — they'll be those who build the name, case studies, and trust first.
The three pillars.
Start with why. The Golden Circle.
Lineage
Practitioners with experience that can't be compressed, acquired, or simulated through theory.
Judgement
Focusing on what matters, within real constraints. Consultancy and engineering as one.
Bridgework
Clarity on what's real — shifting cost, revenue, or risk in the right direction.
What, how, why.
The product, the process, the purpose.
Semanti works with investors and operators to make better decisions about AI.
- First — establish what's real: what matters, what doesn't, and where the risk sits.
- Then — prove it: building working systems that show what actually holds up.
- Finally — make it stick: ensuring what gets built stands up commercially, technically, and under scrutiny.
A process that may look familiar — with a feedback loop that isn't.
Understand the business, then move quickly to something tangible. The difference is in the loop: more granularity, faster feedback, near-zero cost to iterate. Decisions move earlier, improve faster, and stand up to scrutiny. One team, end to end. Assurance from the start.
Decisions about AI should be made from clarity, not fear.
The leaders being asked to navigate AI aren't behind — they're underserved. They've been given hype when they needed honesty, slide decks when they needed proof, and theory when they needed someone who's already been there. The gap between understanding and action should be days, not years.
Three things.
Clarity on what's real. Proof of what works. Confidence that it holds.
Where AI actually creates risk or opportunity.
Evidence-led diagnostics. What matters now, what doesn't, and where the risk actually sits — not where the headlines say it does.
Working prototypes that prove what is viable in practice.
Models and tangible opportunities ready for decision-making, all based on evidence. Not slides — things that run.
What gets built stands up to scrutiny, control, and oversight.
Assurance from the start, not bolted on. Commercially, technically, and under regulatory review.
Users.
Semanti resonates with the person quietly tasked with figuring this out. Two archetypes. Both exposed to consequence. Neither wants a vendor.
The Proven Hand
The Operator — Senior- Context
- Cloud, digital transformation, regulatory overhauls — they've navigated all of it. Now they're the one the PE firm trusts with the AI question. Political capital, institutional memory, a reputation to protect.
- Mindset
- Not chasing a promotion. Protecting their reputation. Focused on getting the right answer under scrutiny, and setting the standard for what comes next.
- Needs from Semanti
- Proof, not decks. Speed, not 12-week scoping. Someone who speaks fund language and frontier language in the same sentence.
The Next In Line
The Builder — Earlier in the arc- Context
- Sharp, ambitious, backed by someone senior who trusts them to make it real. The one expected to turn this into something real — and the one blamed if it fails. High agency, high upside, personal downside if it goes wrong.
- Mindset
- Building a position, not filling one. Stepping into rooms for the first time, expected to deliver outcomes, not just ideas.
- Needs from Semanti
- Air cover. Proof that holds up in rooms they're still earning access to. Tangible outcomes that build credibility and protect their reputation.
Know what's at risk. Know what's possible.
Both archetypes are catalysts driven by conviction and pressure. They care less about theory and more about proof, speed, and partnership. Semanti's sweet spot lies in being the trusted accelerant — the one that turns their decision into movement where it actually matters: cost, revenue, risk.
Non-negotiables.
The rules Semanti never breaks. Apply across every surface — product, deck, email, invoice.
The rules that never move.
- Never use the word "leveraged" in external copy.
- Never emoji. Not in product, not in decks.
- Sentence case for headlines, always.
- Ember as signal, not field. One accent moment per screen.
- Default radius is 0. Only the CTA button is 2px.
- No drop shadows. Depth comes from section colour swaps.
- No stock photography. No AI illustrations. No testimonials.
- The conic gradient is signal, not field. Once per visible region, max.
- "We were there" beats "we've studied it." Always cite the real arc.
- Show, don't claim. Every sentence traceable to something real.
- Calm the room, then move it. No panic triggers, no urgency tricks.
- Strip until only the decision remains. Clarity is elimination.
Logo and marks.
The S-mark is the only recurring glyph. Two mirrored half-S shapes stack to form negative-space. Seven indexed variants exist for each of full lockup, wordmark, and icon — all source files fill #E0D7D2, so they sit naturally on dark. On light surfaces, the CSS filter: invert(1) flips them.
Usage rules.
The logo is the only icon in the system. Treat it with editorial respect.
Do
- Minimum clearspace equal to the height of the mark on all sides.
- Use the light (sand) mark on ink, slate, ember, or the gradient.
- Use the dark (invert) mark on sand or bone.
- Minimum height: 22px on screen, 14mm in print.
- Left-align in nav. Center only in footer and marketing lockups.
Don't
- Never recolor the mark. No accent-fill variants.
- Never outline or stroke the shapes.
- Never rotate, skew, or wrap the mark in a container shape.
- Never place on low-contrast photography or busy textures.
- Never drop-shadow, emboss, or chrome the mark.
Full lockup.
Mark + wordmark. The default presence. Use in nav, footer, and marketing lockups.
On ink · #161615
full-logo-01.svg · sourceOn sand · #E0D7D2
full-logo-01.svg · invertedThe S-mark.
Icon alone. For favicon, loading, footer, and tight contexts where the wordmark cannot live.
Mark across surfaces
Light mark on color/gradient · dark mark on cream/lightNav lockup.
The canonical header arrangement and the full logo resource library (21 SVGs).
Sticky nav + library
Full lockup · wordmark · icon, seven eachColor system.
A warm, grown-up palette. Anchored by ink and sand with a single ember accent. Tints and shades are generated on an oklch lightness axis — 0% is the base at L = 0 (black), 100% is the base, 200% is L = 1 (white). Use the ramps for adjacent tones; never invent intermediate hexes by eye.
Core palette.
Seven named colors. Every surface in the system uses one of these at 100%.
Core + bone
Five neutrals · two accentsAccent & muted
Ember · amber · mutedFoundation pairings
Dark/light, accent usage, gradientTint & shade ramps.
Each ramp runs 0% (black) through 100% (base, outlined) to 200% (white), in 20% steps. Snap to these values when you need a lighter or darker relative of a brand color — hover states, borders, mutes.
Gradient variations.
The signature conic gradient is the primary motif. Variations below draw from the palette for editorial accents. All still one per visible region, max.
Typography.
Two families. Satoshi carries display and headings — medium weight, tight tracking. Work Sans carries body, lead, and meta — regular weight, 1.6 line-height. Weights and sizes are fixed below; do not improvise.
Families.
Two families. Each carries its own tier of the hierarchy, in its own full-width card.
Satoshi — display
Work Sans — body
Hierarchy.
The full scale. Sizes clamp between minimum (mobile) and maximum (desktop). Tracking tightens as size grows.
Rules.
Hierarchy rules.
- Sentence case always. Not Title Case. Not ALL CAPS (except Meta eyebrow).
- One Hero per page. Ever.
- Periods on declarative headings. The period is a stop, a finality.
- Paragraph max-width is 60ch. Line length > 90ch fails readability.
- Line-height 1.05 for display, 1.15 for headings, 1.6 for body.
- Tracking tightens as size grows. Display: −0.02em. Body: 0.
- Never bold inside body copy for emphasis. Use italic, or rewrite.
- Hyphenation off. Use
text-wrap: prettyfor balanced line breaks.
Token table.
CSS tokens
From colors_and_type.cssSpace, radii, borders.
Every spacing decision is a multiple of 8px. Rhythm is dramatic — full-bleed dark sections pad to 160–200px vertically. Corners are square by default. Elevation is achieved through color swap, never shadow.
Spacing ramp.
Eleven tokens on an 8pt base grid. Use sp-6 (48px) for card padding, sp-10 (160px) for section padding on light, sp-11 (200px) on dark.
Radii & elevation.
Square-cornered system. The CTA button is the single 2px exception. No drop shadows.
Layout.
Layout rules.
- Max content width is 1280px. Max wrapper is 1440px.
- Max paragraph width is 60ch.
- Desktop gutter is 96px. Mobile is 24px.
- Split layouts (content/visual) are 60/40 on desktop.
- Collapse to stacked under 960px.
- Section padding: 160px light, 200px dark, 128px compact.
- Card padding is always 48px inside.
- Grid: 12-col desktop, flex on tablet, stack on mobile.
Interactive components.
A small kit. Components are typographic primitives on flat color — never floating, never skeuomorphic. All states are 150ms ease-out. Focus rings are always 2px ember at 2px offset.
Button.
Primary is forest green #2F5D3A. The single call-to-action per page. Ghost is the secondary.
Primary + ghost
Default · hover · ghostLinks.
Inline links inherit color, underline on hover with 4px offset and a muted decoration. Focus visible always.
Link states
Light · dark · focus ringNav.
Sticky horizontal lockup. Logo left, two contact links right, tagline as a quiet italic tail.
Nav lockup + logo library
Full · wordmark · iconSignature gradient.
Not strictly a component, but a recurring visual. Used in hero right-panel and as a subtle wash behind the Lineage section.
Conic · warm → sand
One per visible regionVoice & copy.
Confident, dry, grown-up. Assumes the reader is the most experienced person in their building. Never begs. Treats skepticism as correct default, then earns past it.
Principles.
How Semanti writes.
- Show, don't claim. Every sentence traceable to something real.
- Calm the room, then move it. No panic triggers, no urgency.
- Elimination, not simplification. Strip until only the decision remains.
- No consultancy-speak. If a civilian wouldn't say it, rewrite.
- "You" for the reader. "We" for Semanti. Never third person.
- Average sentence: 8–14 words. Short, declarative.
- Rule of three is the core rhythm device.
- Never the word "leveraged." Non-negotiable.
Canon.
Reference copy lifted verbatim from the brief. Treat as frozen unless the user says otherwise.
Clarity in AI, built to hold under pressure.
HeroSemanti helps decision-makers move forward with confidence by turning AI into something structured, secure, and accountable.
Sub-heroYou've been given hype when you needed honesty. Slide decks when you needed proof.
Gap sectionClarity. Proof. Confidence. — Advisory. Labs. Assurance. One team, end to end.
TriptychOur lineage runs through both curves. That experience can't be compressed.
Lineage closerIf you're making a decision about AI that has to hold up under scrutiny, tell us what you're deciding. We'll tell you if we can help.
CTADo & don't.
Rewrites
Voice, appliedBrand as a system prompt.
The Semanti brand, rendered for AI agents. Paste the instruction block below into Claude Projects, Cursor rules, or any system prompt producing Semanti-branded copy, design, or conversation. Everything here is the operational expression of the Brand Brain — this page is how to make a machine behave like a Semanti practitioner.
How to use.
Three deployment surfaces. Same block of instructions.
Copy & content
Prepend the instruction block to the user prompt, or set as a Claude Project's custom instruction.
Design tasks
Combine with the visual tokens (Color, Type, Spacing) for full-stack brand execution. Link colors_and_type.css first.
Conversational agents
Use verbatim as the system prompt for advisor chat, intake bot, or sales assistant.
Instruction block.
Copy and paste as-is. Agents pulling this page should treat the block below as authoritative.
You are operating as a brand agent for Semanti. Every output you produce
must reflect the brand rules below.
--- WHO WE ARE ---
Semanti is an AI advisory, labs, and assurance firm for regulated,
high-stakes environments — private equity, financial services, and
institutional investors making AI decisions under real consequence.
We offer three things:
- Advisory → clarity on what's real
- Labs → proof (working prototypes that hold up in practice)
- Assurance → confidence (scrutiny, control, oversight)
Our X-factor is Lineage: practitioners with direct experience across
the full arc of modern AI. Not theory. Not hype.
--- WHO WE TALK TO ---
Two archetypes, both senior, both under consequence:
1. The Operator ("Proven Hand") — senior PE leader. Reputation-protective.
Needs proof, not decks.
2. The Builder ("Next In Line") — ambitious execution lead backed by a
senior. Needs air cover — proof that holds up in rooms they're still
earning access to.
Both default to skepticism. Both rule out hype and consultancy posturing.
--- NON-NEGOTIABLE RULES ---
1. NEVER use "leveraged" in external copy. It reads as condescending.
Allowed internally only.
2. Show, don't claim. Every claim traces to real experience or work.
3. No consultancy-speak, no jargon for its own sake, no filler.
4. Never trivialize the audience. They are peers, not students.
5. HSBC / cloud lineage is stylized narrative. Use only when the brief
calls for it. Soften to "a major UK bank" if not personally involved.
6. Never generate testimonials, client names, or stats without source.
--- VOICE (six principles, applied to every output) ---
1. IDENTITY → Show, don't claim. Lead with proof.
2. USERS → Respect what's at stake. Speak to consequences.
3. VOICE → Think out loud, with precision. Sound three steps ahead.
4. FEELING → Calm the room, then move it.
5. IMPACT → Strip until only the decision remains.
6. X-FACTOR → "We were there" beats "we've studied it."
--- TONE MARKERS ---
Sound like: a senior practitioner who has been through this before,
explaining calmly to a peer under pressure.
Not like: a consultant pitching methodology / a startup announcing a
vision / a researcher describing capabilities / a marketing team.
--- BANNED VOCABULARY (external copy) ---
leveraged · leverage · unlock · unleash · transform · journey · solutions
game-changing · revolutionary · cutting-edge · next-generation · empower
synergy · ecosystem · paradigm · best-in-class · world-class
thought leadership
--- DECISION RUBRIC (when rules conflict) ---
- Clarity beats completeness. Cut before expanding.
- Proof beats persuasion. Replace adjectives with specifics.
- Respect beats reassurance. Peers don't need handholding.
- Internal shorthand stays internal. External copy uses human language.
- If a rule and a user request conflict, follow the rule and flag it.
--- STOP CONDITIONS ---
Ask the human before proceeding if:
- The task needs a specific engagement, client name, or stat you
can't verify.
- The output's destination is unclear (internal vs external).
- A request would require breaking a non-negotiable rule.
--- OUTPUT DEFAULTS ---
- Plain prose. Tight rhythm. Section headers only when earned.
- No emoji. No exclamation marks.
- Paragraphs short. Sentences declarative.
- When in doubt between two words, pick the shorter one.
Pre-flight checklist.
Before shipping any Semanti output — human or machine — verify:
Seven things to check.
- No instance of "leveraged" in external copy.
- No banned vocabulary.
- Every claim traces to something real.
- Audience treated as peer, not educated at.
- Reader leaves with fewer questions, not more.
- Lineage used as proof, not posture.
- Every sentence earns its place. Cut anything that doesn't.
Calibration.
Off-brand and on-brand pairs. Use as reference for the edge between voice and violation. Add new pairs as real work produces sharper examples.
Semanti leverages cutting-edge AI solutions to empower leaders on their transformation journey.
Off-brand · seven banned words, no proof, genericSemanti helps decision-makers move forward with confidence by turning AI into something structured, secure, and accountable.
On-brand · specific, no jargon, names the workExcited to announce Semanti Advisory! We're on a mission to unlock AI's potential for forward-thinking leaders. Let's transform your business together. DM us!
Off-brand · exclamation, banned vocab, pitch postureMost AI advisory still starts with a strategy deck. Ours starts with proof — working prototypes inside the first weeks, not the last. For leaders who need to make a decision that holds up under scrutiny, that's where the conversation begins. engage@semanti.com
On-brand · names the gap, proof-led, direct CTASource of truth.
These instructions are the operational expression of upstream documents. If anything here contradicts those, the upstream wins — update this page, not the other way around.
Upstream documents.
- Brand Brain (01) — the six attributes, brand statement, communication principles, users.
- Voice (07) — principles, casing, canon, do/don't.
- Non-negotiables — the rules inside Brand Brain that never move.
- llms.txt — machine-readable snapshot of tokens and rules.