← All case studies + AI/ML · Live since 2024 · quikest.ai
Quikest.ai logo Client engagement · Quikest.ai
Case · 42 · AI/ML

AI persona workspace for product teams

The workspace where the whole team points at the same person.

A collaborative AI workspace for marketing, product, and sales teams. Generate personas in minutes, interview them like real customers, and export them into a codebase when the work moves downstream.

Personas per product definition
3–5
AI-generated in a single conversation
LLMs orchestrated
4
Gemini · GPT-4o-mini · Imagen · Flash Image
Export formats
3
.claude/ · .cursor/ · generic ZIP for agent context
Real-time cursors
1 per teammate
Y.js CRDT · Hocuspocus WebSocket sync
Quikest.ai marketing hero.
We built Quikest because we kept watching the same failure mode. A marketing lead, a PM, and a sales engineer would each say the customer in the same meeting and each mean someone different. The disagreement about the headline was, underneath, a disagreement about who was reading it. Quikest is what we built to end that argument.
The founder · Quikest.ai
Scope

What we were hired to build.

Quikest.ai is an AI-powered collaborative workspace for product teams to synthesise customer personas and anchor strategy around a single source of truth. It serves product managers, UX researchers, founders and developers who need to turn a product definition into actionable persona profiles — with AI-generated persona synthesis, rich-text biography editing, a chat interface to “interview” personas, and real-time multi-user collaboration.

How we planned

A two-stage AI pipeline.

We designed a two-stage AI pipeline: URL scan or manual product definition feeds into Google Gemini 3 Flash for rapid persona recommendation, then OpenAI GPT-4o-mini synthesises the full persona profile with demographics, psychographics and biography. A fine-tuned MedGemma model generates contextualised profile images. The workspace architecture uses TipTap v2 for rich-text editing, Y.js CRDT for offline-first multi-user editing and Hocuspocus for WebSocket sync. Auth is Clerk plus Supabase RLS; billing is Stripe with Free and Pro tiers.

How we delivered

A live SaaS with agent-context handoff.

Quikest launched as a live SaaS with a 7-day free trial and a public demo path. Every workspace gets multi-user collaboration out of the box — team members real-time-edit persona biographies, comment with resolve states, and access shared document libraries. The “Ask AI” chat surface lets teams interview personas as if they were real customers. Export ships personas straight into Claude Code, Cursor and GitHub Copilot via agent-context ZIP files.

Quikest workspace dashboard — projects view.
Workspace dashboard — every workspace is a shared source of truth.
Sample persona 1 — founder-office context.
A fine-tuned MedGemma generates contextualised profile images.
Persona biography longform editor.
TipTap v2 rich-text editor with real-time CRDT collaboration.
Ask AI sidebar — interview a persona as if they were a real customer.
Ask AI — interview the persona as if they were a real customer.
Sample persona 2 — designer at desk.
The designer — face-consistent portrait via Gemini 2.5 Flash Image.
Sample persona 3 — traveler in an airport.
The traveler — setting-aware 16:9 environment via Gemini Imagen 4.0.
Templates library.
Templates library — every workspace starts from a shared starting point.
Pricing surface — Free and Pro tiers.
Free and Pro tiers via Stripe with a 7-day free trial.
What we learnt

Ownership converts adoption.

Persona generation is fastest when driven by a single conversation with the AI, not a multi-step refinement loop. Teams trust AI-generated personas more when they can edit and own them immediately — ownership converts adoption. The real value isn’t AI speed; it’s making personas a living collaborative document rather than a one-off deliverable in a Google Drive folder.

How the client improved

The discovery motion resonates more than the persona card.

Quikest supports multi-user workflows across startups and mid-market product teams, with the persona-chat surface driving primary engagement — indicating that the discovery motion resonates more than the persona card itself.

01 / 02 · Client CEO testimonial
On brief, on time, and somehow still beyond what we imagined.
Aaron Kettl
CEO, Quikest · 2024
02 / 02 · Early-access user · 2024
The persona started to feel less like a document and more like a colleague we could tap on the shoulder.
Early-access user
Ask-AI feedback · Quikest · 2024
Stack · in receipts

The engineering, in one page.

Editor stack
TipTap v2
Rich-text with real-time CRDT collaboration
Persona images
MedGemma
Fine-tuned model generates contextualised profile images
Auth + billing
Clerk + Stripe
Supabase Row-Level Security · Free and Pro tiers
Primary engagement
Ask AI chat
Teams interview personas as if they were real customers

A two-stage pipeline (Gemini 3 Flash → GPT-4o-mini) synthesises full persona profiles from a URL or a paragraph. Real-time multi-user editing on TipTap v2 + Y.js CRDT + Hocuspocus. Fine-tuned MedGemma generates profile images. Auth is Clerk plus Supabase RLS. Billing is Stripe on a Free + Pro model with a 7-day free trial. Export ships straight into Claude Code, Cursor and GitHub Copilot via agent-context ZIP files.

Next case

AIHumanize.tech — the humanization tool that skips the arms race.

One pass. Eight registers. Sentence-level control. Built by people who studied how detectors work — and fine-tuned our own model so you don’t have to think about them again.

Have something worth building?

You THINK, We BUILD.

Send a paragraph about what you’re building to hello@creative-mantra.com. We reply within 24 hours with next steps and, for qualified scopes, a fee letter within four working days.