Skip to content

Update landing page: trust-first redesign, pricing & about pages, nav#14

Open
jonnyholland wants to merge 49 commits into
mainfrom
eng-220-update-landing-page
Open

Update landing page: trust-first redesign, pricing & about pages, nav#14
jonnyholland wants to merge 49 commits into
mainfrom
eng-220-update-landing-page

Conversation

@jonnyholland

@jonnyholland jonnyholland commented Jun 30, 2026

Copy link
Copy Markdown
Contributor

Rebuilds the marketing site around a clearer, trust-first message and brings the landing, pricing, and about pages in line with a cleaner Linear-style design.

Landing page

  • Nine Figma-aligned sections: static hero with a live in-app ProductWindow simulation, a four-pillar overview (Trust / Security / Governance / Ownership), how-it-works, dedicated Trust/Security/Governance/Ownership/Intelligence sections, and a redesigned footer.
  • Placeholder imagery replaced with in-product proof panels (score bars, encryption/policy status cards).
  • "Powered by DAIS" section: an animated diagram showing ButteryAI wrapping DAIS, which runs the request through five narrated steps (screen & govern → understand → plan → run across nodes in parallel → evaluate), with node internals (own intelligence + LLM), an encryption/audit blanket, advantage cards, and a feedback loop.
  • Shared design tokens + section mixins; reduced type scale, tightened section spacing.

Pricing page (/pricing)

  • Built from the server's PlanCatalog tiers (Free / Pro / Team / Enterprise) — no invented numbers.
  • Linear-style: line-separated plan columns with bottom-aligned CTAs, a subtle founder note, and a checkmarked, sticky-header comparison matrix sourced from PRICING_TIERS.md.

About page (/about)

  • Linear-style: "Building AI you can trust — and own." title, one-sentence USP intro, the reused ProductWindow simulation, a product-info section, and the team grid.

Shared / infra

  • Floating icon NavBar (Home / Pricing / About) with active state, across all marketing pages.
  • Reusable ProductWindow matched to the real app chat UI.
  • Per-page tab titles via a useDocumentTitle hook.
  • Enterprise CTAs open a prefilled sales email; waitlist posts only to the ButteryAI server (Firebase write path removed).

Notes

  • The 8 retired home sections are no longer rendered; files remain in place (not deleted).
  • The unrelated pre-existing Login/Form/Form.tsx WIP is intentionally excluded.

🤖 Generated with Claude Code

jonnyholland and others added 30 commits April 6, 2026 16:32
Aligns with the native app's new auto-title flow where "New Chat" is
a brief placeholder before the DAIS suggests a real title.

Co-Authored-By: Claude Opus 4.6 (1M context) <[email protected]>
Restructure the home page into nine Figma-aligned sections: a new static
hero with a live product window, a four-pillar overview, how-it-works,
and dedicated Trust, Security, Governance, Ownership, and Intelligence
sections, plus a redesigned footer. Replace placeholder imagery with
in-product proof panels (score bars, encryption/policy status cards) and
add shared design tokens and section mixins.

Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>
Introduce a /pricing page mirroring the server PlanCatalog tiers (Free,
Pro, Team, Enterprise) and an /about page reusing the team section, both
with a shared top nav and footer links. Drop the Firebase write path from
the waitlist form so submissions go solely through the ButteryAI server
waitlist endpoints.

Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>
Slim the plan cards to price, aligned CTA, and a short highlight list,
and move the full per-tier capability matrix into a grouped comparison
table below. Give the price block a fixed height so every call-to-action
lines up across cards.

Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>
Replace the boxed plan cards with Linear-style columns divided by thin
lines, a subtle tint on the popular plan instead of a border, and CTAs
pinned to the bottom so they align across all four plans.

Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>
Prefix every included capability with a green check and mark absent ones
with a muted ✕, mirroring Linear's scannable matrix. Make the plan header
row sticky while scrolling and add a subtle row hover.

Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>
Drop the Most-popular badge and blue highlighting from the Pro column
(cards and comparison table), shrink the price, and present the founder
discount as a subtle chip. Rework the page header into a clean,
left-aligned Linear-style title with the founder note inline.

Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>
Shrink the plan prices and page heading, drop the Enterprise "starting
at" sub-line and the founder chip under each price, and restore the Pro
plan's primary call-to-action button.

Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>
Scale down the shared footer call-to-action typography so it no longer
dominates the page across home, pricing, and about.

Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>
Shrink the hero, section headers, split-section titles/body, and card
headings across the home page so the typography matches the cleaner,
smaller scale used on the pricing page.

Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>
Introduce a shared, fixed pill-style nav bar (Home/Pricing/About icons
with active state) and use it across the marketing pages in place of the
per-page top bars. Left-align the about header to match the pricing page
and shrink the team heading to the same scale.

Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>
Increase the home hero's top padding so the logo clears the floating nav
with the same gap it has to the title, and replace the pricing nav icon
with a symmetric, centered dollar sign.

Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>
Replace the boxy dollar glyph with a smooth, glyph-like S and a centered
full-height bar so it reads as a proper dollar sign.

Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>
Recenter the S vertically and shorten the bar so it overhangs equally at
the top and bottom.

Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>
Rebuild the about page Linear-style: drop the eyebrow label, lead with a
"Building AI you can trust" title and a one-sentence USP intro, add the
DAIS hive visual, and a product-info section below. Left-align the team
heading. Add a useDocumentTitle hook and set tab titles across home,
pricing, and about, and remove the now-redundant on-page "Pricing" label.

Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>
Bring back the on-page "Pricing" heading and give the about hero, info,
and team sections a shared content width so their titles line up on the
same left edge.

Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>
Extract the hero's product-window mock into a reusable ProductWindow
component and render it on the about page in place of the hive image.
Give each placement a distinct, realistic example: a Customer Support
cluster on the landing hero and a Sales Assistant cluster on about.

Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>
Update ProductWindow to mirror the in-app chat: a centered conversation
header with side rules and info icon, a reset control, a tailed blue user
bubble, a gray AI bubble, and a compact tokens/time/accuracy metrics row.

Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>
Bring back the app window title bar with the cluster name + Local
selector, and use the centered section divider for a separate
conversation title (Q3 Refund Policy Inquiry on the hero, Enterprise
Plan Inquiry on about).

Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>
Drop the bottom metrics (tokens, time, accuracy) from bold to a normal
weight in both placements.

Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>
Anchor the section titles, subtitles, and supporting lines to the left
edge so they line up with the cards below instead of being centered.

Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>
Route the hero "For Enterprise", footer "Talk to us about enterprise",
and pricing "Talk to sales" buttons through a shared helper that opens
the mail client with a prefilled subject and inquiry template.

Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>
Call out the proprietary DAIS engines (Routing, Management,
KnowledgeIntelligence, TrustIntelligence, and epistemic verification) in
a new section after how-it-works, each tagged with its internal name.

Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>
Replace the engines list with a pipeline diagram — prompt routes through
the Routing and Management engines, gets scored by the evaluation engines
(Knowledge/Trust/Epistemic), and returns a trusted answer — using the
brand hexagon motif and a feedback-loop caption.

Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>
Show the engines running inside a labeled DAIS boundary: inputs on the
left, a request flowing through a security/trust gate into the honeycomb
Cortex + nodes (with the five engines), out through output evaluation and
encryption, to outputs on the right. Add a subtle animated packet and a
feedback-loop caption; respect reduced-motion.

Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>
Reframe the section as a two-layer diagram: ButteryAI (the product) sits
on top, powered by DAIS (the distributed AI system) below. Inside DAIS,
arrange the engines as a connected left-to-right flow — Security & Trust,
Routing, Management, Evaluation (Knowledge/Trust/Epistemic), Encrypt &
audit — with a packet animating through, so the connections and flow are
clear.

Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>
Restructure the engines section vertically: Request enters ButteryAI,
which wraps DAIS, which contains the engines as a top-to-bottom timeline
(Security & Trust, Routing, Management, Evaluation, Encrypt & audit) with
a packet flowing down, ending in a Trusted output. Nesting makes the
ButteryAI-runs-on-DAIS relationship explicit and removes horizontal
scroll.

Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>
Insert a Node processing step between Management and Evaluation, shown as
a honeycomb hexagon with a nested "inside each node" panel: each node has
its own intelligence, knowledge, and context, and runs an LLM (local GGUF
or remote OpenAI/Anthropic).

Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>
Introduce a DAIS Core header inside the DAIS box — the orchestrator that
holds and runs every engine in-process, with pure-Swift/in-process/no
token loss efficiency chips — so it reads as a coordinated system rather
than a loose pipeline.

Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>
jonnyholland and others added 15 commits July 1, 2026 12:50
Rework the engine timeline into five numbered steps that describe what
happens (screen, understand & route to nodes, plan, run across nodes in
parallel, evaluate & rerun) instead of listing engines. Call out
multi-node parallelism and move encryption/audit into a cross-cutting
band that wraps every step rather than a sequential step.

Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>
Rename the "Epistemic" evaluation chip to "Confidence" and add Gemini to
the remote model examples.

Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>
Revert the diagram to centered, drop the DAIS Core band, and add a
checkmark advantages column to the left of the diagram (no token loss,
end-to-end encryption, tamper-evident audit trail, fast & efficient,
etc.) to declutter the container and surface the benefits.

Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>
Drop the request/output pills from the DAIS diagram, move the advantages
into cards above it and the feedback note into a card below. Reduce card
font sizes ~20% across the foundation, how-it-works, and intelligence
sections and tighten card padding so they read less stretched.

Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>
Fold policy governance into the first DAIS step (screen & govern) and add
Governance & policy, Model agnostic, Custom evaluations, and Auto-learning
to the advantage cards.

Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>
Rename step 2 to "Understand the request" with a lightbulb icon, reshape
the node badge to a rounded square (matching the others) using the
network icon, and center the section title and subtitle.

Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>
Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>
Replace the awkward inline LLM label with a small "inside each node" card
showing two layers — the node's own intelligence engines and the LLM it
runs — each with an icon and description.

Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>
Reduce the shared section padding (160/120/80 -> 112/88/64) so sections
no longer feel top-heavy with excess empty space now that the card type
is smaller.

Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>
Group the title, subtitle, and note together and reduce the gap to the
step cards so the header reads as one cohesive block.

Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>
Bump the check-item gap (16 -> 18) on the trust, security, and governance
sections so two-line items are less cramped.

Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>
Tighten the intelligence and foundation section headers (subtitle/lede
gaps and header-to-cards margin) to the same rhythm as how-it-works.

Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>
Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>
@jonnyholland jonnyholland force-pushed the eng-220-update-landing-page branch from df2221e to 8847690 Compare July 1, 2026 21:56
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant