AnyTeam: AI-native sales startup

Taking a startup website from 0 to 3,500 visitors

Role
Designer and Developer
Timeline
Continous
Tools
Framer, Figma, Claude
Domain
B2B · Sales Startup

Context

AI-native sales startup. $10M seed. Invite-only product. Stealth to beta

I built AnyTeam's entire web presence from scratch, working within an existing design system to ship before launch day. The site needed to serve investors evaluating credibility and sales professionals deciding whether to request beta access.

I owned the process from concept to ship, using my technical background and AI tools to bridge design and engineering.

Results

3,500

Active users from a standing start

1 week

Completed one week before launch

0 → 1

Brand taken from stealth to live

Discovery

What I learned before opening Figma

Stakeholder Interviews

Founders feared looking generic.

Creative direction was "retro coastal meets liquid glass." The deeper need: the site had to signal taste. In a market of identical AI pages, intentionality was necessary.

User Behavior

AEs scan pages in 30–60 seconds.

Fixed sequence: "What is this?" → "Is it legit?" → "How do I get in?" Any friction = bounce. This became the IA blueprint.

Information Architecture

A narrative IA

A hero CTA on an invite-only product is presumptuous. Burying the form loses conversions. Solution: build trust progressively so the signup feels earned.

Homepage User Flow

"What is this?"

Lead with the value proposition. Answer the first question every visitor has.

Key Decisions

Design decisions

UX

Show users what their day looks like with AnyTeam instead of listing capabilities.

Brand

Visual-heavy, interactive components differentiate from the "just another AI startup" pattern.

Prioritized

Trust through familiarity. Similar structure with slight variations lets users build confidence as they navigate.

Traded off

Visual differentiation between features. At this stage, coherence mattered more than variety.

UX

Consistency between product and marketing reduces cognitive gap. The site feels like the product before users touch it.

Process

4-week timeline. Working within the system freed focus for narrative, hierarchy, and motion.

Challenges

Problems I solved along the way

Video

Product demos didn't translate well to mobile. Removed video from mobile since our primary audience browses on desktop. Polished on one platform beats compromised on both.

Technical

Google Analytics, domain connection, Mailchimp integration. Not design problems, but real blockers I had to solve quickly.

CMS

Blog required Framer's built-in CMS. Designing templates flexible enough for varied content while maintaining visual consistency was a new challenge.

Outcomes

The impact

0

Active users

From a standing start

0s

Avg. engagement

vs. 15–30s benchmark

0 week early

Delivered

Before launch date

Reflection

Bridging design and engineering

Working at a seed-stage startup means wearing many hats and moving fast, two things I genuinely enjoy. Collaborating with a driven team and turning ideas into real things quickly is where I thrive. Designers are increasingly owning more of the engineering process, and my technical background let me close that gap naturally. I used AI tools like Claude and Framer for rapid iteration and execution, but every decision was still guided by human judgment and a close eye on the details.

Saiyushi Kumar

Made with love and iced lattes

Saiyushi Kumar

Made with love and iced lattes

Saiyushi Kumar

Made with love and iced lattes