SaaS & Cloud Platforms - 2023–2025

SAIT.ai

An AI platform that turns a CV, a content plan, a slide deck, and a website into one login instead of four separate tools.

Role
Research, UX/UI design
Platform
Web app
Scope
CV generator, content planning, presentations, graphics, website builder
Year
2023–2025
Diagram from the SAIT project showing one account connected to CV generation, content planning, presentations, graphics, and a website builder, next to the SAIT wordmark.
The whole product in one diagram: one account (SAIT.ai) sitting between the user and six generators - CV, invest plan, content plan, presentations, graphics, and website builder.

The problem

Five outputs that don't share a shape

SAIT set out to put CV writing, content planning, presentation design, graphics, and website building behind one AI account. Each generator produces a completely different kind of thing - a resume, a content calendar, a slide deck, a live website - for a different kind of user: a job seeker, a marketer, a founder with no design background. Getting all of them to feel like one product, instead of four demos stitched together, was the actual design problem.

CV Generator

A résumé built from a few structured fields

The user enters full name, years of experience, area of expertise, and skills, then adds work-experience entries one at a time - company, dates, title, and a short description. The panel on the right stays visible while filling the form and explains, in plain language, how the output will be built: templates chosen for the industry, formatting kept consistent, and keywords worked in that applicant-tracking systems tend to filter on.

CV Generator form asking for full name, expertise, years of experience, and skills, with work-experience entries below and an AI info panel on the right.
CV Generator workspace - name, expertise, experience, skills, and a running list of past roles on the left; on the right, a panel explaining how the AI tailors templates and keywords to the profile as it's filled in.

Content Plan

A publishing calendar from a goal and a budget

Instead of a blank calendar, the user answers six short prompts: what the business does, what the goal is, who the audience is, how often to publish, what tone the content should take, and a weekly budget. "Generate Content Plan" turns those six answers into a schedule - the same information a strategist would ask for in a briefing call, just typed into fields instead of said out loud.

Content Plan generator form asking for business details, goal, audience, publishing schedule, content wishes, and budget, shown on a laptop mockup.
Content Plan generator - business details, goal, audience, a publishing cadence ("once a week"), tone notes, and a weekly budget feed a single "Generate Content Plan" action.

Presentations

A slide count, a topic, and a starting structure

The presentation tool asks for two things up front - how many slides (a range like 5–10) and a description of the topic - then hands over a choice of three visual structures before anything gets generated: a dark text-and-image layout, a warm gradient layout, and a photo-led layout with an overlaid heading. Picking a structure before generation means the AI is filling a shape the user already approved, not guessing at one.

Presentation generator form asking for slide count and topic, with a choice of three slide structure templates shown as thumbnails.
Presentation generator - slide count and topic on top, three selectable slide structures below; "Generate Presentation" only becomes the next step once one is chosen.

Approach

One shape, reused five times

I designed a single generation pattern and used it for every tool: describe what you need in a short form, choose from AI-suggested structures, get a result you can still edit. A recruiter filling in CV fields and a founder describing their business for a website are answering completely different questions, but both move through the same three steps to get there.

Learn the flow once. Every generator on the platform follows it.

Website builder

From a business description to a live draft

This is the pattern at its fullest. The user enters business type, company name, address, up to four core services, and a free-text description of what the company does, then picks a color theme and uploads a logo. "Generate Website" produces a full multi-page site - hero, services, about, contact form - laid out in the chosen theme, shown at both desktop and mobile widths before anything is final.

Website builder flow across several screens: the business-details form, a theme picker, and generated page previews on desktop and mobile.
The full input-to-preview sequence: business details and services on the left, a color-theme picker, and the generated pages previewed across breakpoints before the user commits to a result.
Generated website result screen with options to download the site as a zip, edit it, or preview it, next to an AI-produced page layout.
The result screen. Three actions sit at the top - Preview, Edit Website, Download Website (.zip) - so the generated site is never a dead end: take the file, keep editing here, or both.

Plans

Bundling by who's using it, not by feature count

Four tiers, split by role rather than by how many tools are unlocked: Work ($60/month) covers the CV Generator and an invest plan; Study ($80/month) covers subjects, presentation help, and an "explain me" tool; Business ($159/month) covers presentations, websites, a menu generator, interior design, marketing and content plans, post writing, and image generation; "Open all AI" ($259/month) unlocks everything. Each plan gets a seven-day free trial except the top tier, which is a straight purchase.

SAIT pricing screen split into Work, Study, Business, and Open all AI tiers, each listing a different set of included generators.
Plan selection - Work and Study bundle a couple of relevant tools each; Business bundles eight; Open all AI removes the bundling question entirely for a flat price.

Process

How the platform got built

  1. Research

    Mapped where AI could remove friction across CV writing, content planning, presentations, and site-building for different kinds of users.

  2. Shared architecture

    Built a modular system so five very different generation flows could share one interface language and one codebase underneath.

  3. Generator by generator

    Designed and shipped each tool - CV, content plan, presentations, graphics, website builder - as its own module inside the same shell.

  4. Hardening

    Tuned the AI compute pipeline and strengthened data security once every generator was live and carrying real usage.

Design decisions

What actually shaped this

Outcome

One account, five generators, live

SAIT shipped as a working multi-surface platform - CV generation, content planning, presentation design, graphics, and a website builder, all under one account and one interface system. The shared pattern was refined with each generator that went live, so the fifth tool cost far less to design than the first.