11 Days to Build.
Seconds to Connect.
AI
Quality Assurance
Frontend
Backend
Project Management
Projects
Snapshot
A rapid prototyping engagement to validate a new video-based dating experience.
Client
Early-stage startup (social/dating)
Engagement
Rapid Prototyping Project (RPP)
Product
Web-based video speed-dating app
Timeline
11 days (kick-off → demo)
Outcome
A fully functional PoC covering onboarding, events, matching, timed video sessions, messaging, and admin. All delivered in just 11 days and ready for live testing.
What Defined the Delivery
We focused on delivering a real, end-to-end product within a fixed timeline, using AI-assisted development to move fast while keeping the system stable enough for real users.
Tech & Tooling
Povio monorepo • Product prototype tool connected to our Povio React Template • Povio Admin Template • Robodev (in-house backend AI code generator) • Daily.co for video
Challenges
Solved
Tight timelines. Evolving requirements. High expectations. We approached each challenge with a clear focus. Build fast, stay lean, and keep everything testable and scalable.
Designing a Product With No Requirements
Challenge
The client had only a rough concept and an unvalidated prototype.
Solution
We reframed the goal: build just enough real product to run a true experiment. We defined flows around onboarding → events → matching → calls → connections, keeping everything lean and testable.
AI-Accelerated Backend Without Losing Control
Challenge
Generate backend services quickly without creating unmaintainable code.
Solution
We defined OpenAPI contracts for users, events, matches, sessions, and messaging. These contracts powered Robodev, producing the initial services that engineers refined and validated.
QA at Startup Speed
Challenge
Traditional QA cycles wouldn’t fit the timeline.
Solution
QA joined mid-sprint as a collaborative partner. Testing onboarding, event logic, reconnections, permissions, and call transitions in tight feedback loops.
Building End-to-End in Just 11 Days
Challenge
Deliver a functional system: auth → video → messaging → admin—under extreme time constraints.
Solution
We leaned heavily on Povio’s internal tooling:
-Product prototype tool connected to our Povio React Template for instant scaffolding
-Povio monorepo conventions
-Admin template for operational control
These eliminated boilerplate so engineers could focus on logic and polish.
Delivering a Smooth Video Experience
Challenge
Video calls needed to feel effortless, paced, and intuitive.
Solution
-Integrated Daily.co for video
-Added our own timing + session logic
-Surfaced shared interests + icebreakers
-Built a clear transition system (‘extend & end’)
By Day 9, the core matching + calling experience was fully functional.
Our Process
11 Working Days
Day 1
From Prototype to Frontend Foundation (In Parallel With Architecture Setup)
Starting from a rough Figma Make prototype, the team scaffolded the web app using the Povio React Template while frontend and AI engineers built the UI foundation and, in parallel, the backend engineer and project team designed the system architecture to support upcoming product flows.
Day 2
Early Flows Take Shape & Contract Design Begins
The frontend team generated the initial structure: screens, routing, navigation, and shared components, giving the app its first recognizable shape. In parallel, the backend team drafted the OpenAPI specifications for users, events, matches, sessions, and messaging. These definitions would soon become the source of truth for backend generation.
Day 3-4
From Solidified Contracts to Generated Backend Services
By this stage, the frontend already resembled the emerging product, while the backend team finalized robust OpenAPI contracts to drive large-scale code generation, and our AI tooling engineer used Povio’s in-house generator to produce the first backend services and endpoints, giving the team a runnable foundation aligned with the intended architecture.
Day 4-5
Engineering Review, Integration Points & Refinement
Engineers reviewed and cleaned up the generated code, wired integrations, validated behavior, and ensured all services matched the contract-defined expectations.
Day 6-7
Frontend + Backend Integration & Operational Tooling
With both sides evolving in parallel, the team connected the frontend and backend to get real data flowing through the app. We also introduced an admin dashboard using Povio’s internal template, enabling the client to create events, monitor activity, and adjust parameters without developer involvement.
Day 8-9
Bringing Video Speed-Dating to Life
To deliver the core 1:1 video dating experience, we integrated Daily.co across frontend and backend with room creation, permissions, session logic, timed rounds, in-call icebreakers, and clear transitions, enabling by Day 9 a complete end-to-end flow from joining an event to matching, dating, and connecting.
Day 10
QA as a Partner
With the core paths in place, QA joined the loop:
validating onboarding, event joins, matching, and calls
stressing reconnections, permissions, and transitions
collaborating in rapid cycles with frontend and backend for fast fixes.
Day 11
Polish & Demo Readiness
To deliver short, meaningful 1:1 conversations, we integrated Daily.co across frontend and backend with rooms, tokens, permissions, session logic, timed rounds, in-call icebreakers, and clear transitions, enabling by Day 9 a complete end-to-end flow from joining an event to matching, dating, and connecting.
What We
Delivered
The result was a fully functional product, not a demo. Within 11 working days, Povio delivered all the core building blocks required to launch, operate, and evaluate a video speed-dating experience.
Auth & Onboarding
Google sign-in, 18+ gate, profile details, predefined interests.
Post-Call Logic
Feedback, connection opt-ins, simple 1:1 messaging with link blocking.
Events & Matching
Upcoming/live events, waiting room, AV checks, auto-matching into 1:1 sessions.
Admin Dashboard
Client-side operational control of events, users, and system activity.
Timed Video Dates
Daily.co video calls with timers, optional extensions, shared interests, icebreakers, and clear controls.
What We
Learned
Ideas that would normally die in discovery can be tested in weeks, not months.
AI tools don’t replace engineers—they remove the boilerplate, freeing engineers to focus on architecture, UX, and quality.
Our internal monorepo, templates, and backend code generator provide real leverage when timelines and budgets are tight.
In 11 days a fuzzy idea became a real product ready for real users.
“AI-assisted tools like Bolt and Robodev allowed us to move incredibly fast while maintaining strong code quality. What would normally take weeks was delivered in a fraction of the time.
Tadej Rebernjak
Product Prototyping Team @ Povio
.avif)
.avif)
.avif)
.avif)
.avif)
.avif)
Next Case Study
Pump Club
UI
Backend
UX
DevOps
Quality Assurance
Frontend
Project Management