PM Accelerator Internship

Director's Brief

The Context

PM Accelerator 11-week AI product development internship. MyHandyAI is an AI-powered home repair assistant helping non-experts complete DIY projects safely through step-by-step visual guidance.

My Role

Product Design Intern working alongside 3 designers, 4 product managers, and an engineering team throughout the full product development cycle.

My Contributions

I designed step-by-step process information architecture used across all repair scenarios, created AI chat feedback system with error recovery flows, built e-commerce integration patterns for tool selection and purchasing, and contributed to design system standardization.

Program

PM Accelerator Internship

Duration

11 Weeks (June-September 2025)

Team Structure

3 Designers • 4 PMs • Engineering Team

My Responsibilities

User Flow Design

Wireframing & Prototyping

Design System Application

Developer Documentation

Tools Used

Figma Discord

AI-Specific UX

Designing for hallucination, error recovery, and user trust

Cross-Functional Collaboration

Working with PMs, engineers, and external mentors

Design Systems

Applying and documenting consistent patterns

Technical Documentation

Creating developer handoff specifications

SCENE 01

Step-by-Step Process Design

Information Architecture & Navigation System

Before My Contribution

After My Design

What I Designed

Complete information architecture for step pages including Tools Needed, Time Estimate, Instructions, Safety warnings, and Tips. I standardized how AI guidance is presented across all scenarios—a design system decision affecting the entire product.

SCENE 02

AI Chat Feedback & Error Recovery

Interaction Design for AI Reliability

Initial Chat Interface

With Feedback Mechanism

Error Recovery Flow

Quick Response Chips

What I Designed

Thumbs up/down interaction flows, response regeneration functionality, 4 distinct conversation recovery paths, and quick response chips for conversation scaffolding. This solved a critical AI UX problem: what happens when AI fails.

SCENE 03

E-Commerce Integration UX

Connecting AI Guidance to Purchase Decisions

Before Expert Feedback

After My Redesign

What I Designed

Tools/materials selection page with required/optional badges, Amazon purchase flow integration, cost tracking functionality, and selection states with inventory management patterns. I connected AI guidance to real purchase decisions, a complex interaction design challenge.

SCENE 04

Wireframe Evolution

From Exploration to Implementation

Low-Fidelity Exploration

Early exploration phase focusing on core functionality and user flow structure

Mid-Fidelity Standardization

Refined interactions with consistent patterns and design system application

High-Fidelity Polished Version

Final polished version ready for developer handoff and implementation

SCENE 05

Design System & Documentation

Systematic Thinking Beyond Screens

Component Library & Developer Documentation

Comprehensive design specifications including reusable patterns, component standards, and technical documentation enabling seamless engineering implementation. My CS background helped bridge design-development communication.

Key Learnings

Growth Through Real-World Experience

What I Learned

AI Will Fail—Design for Failure

AI systems aren't perfect. Feedback loops must be obvious and valuable. Transparency builds trust more than perfect accuracy ever could.

Collaboration Shapes Better Design

Design decisions directly impact engineering complexity. External expert feedback reveals internal blind spots. Documentation quality determines implementation success.

Personal Growth Through Initiative

My CS background helped bridge design-development communication. I learned to lead design decisions while staying collaborative. Initiative matters—proactive problem-solving got me chosen for this team.

Program Outcome

Delivered

Complete step-by-step UX patterns implemented in team wireframes
AI feedback system ready for developer handoff
Design documentation enabling engineering implementation

Impact

Designs validated by industry expert with Tesla/Walmart background
Presented at Demo Day showcasing full product journey
Contributed to team's successful 11-week sprint completion

Recognition

Successfully completed PM Accelerator's intensive 11-week AI product development program with certification and recommendation letter.

View Certificate

Live Demo

Experience the product we built during the internship. The MVP is live and available for testing.

Try MyHandyAI

PM ACCELERATOR

Director's Brief

The Context

My Role

My Contributions

Program

Duration

Team Structure

My Responsibilities

Tools Used

AI-Specific UX

Cross-Functional Collaboration

Design Systems

Technical Documentation

Step-by-Step Process Design

What I Designed

AI Chat Feedback & Error Recovery

What I Designed

E-Commerce Integration UX

What I Designed

Wireframe Evolution

Low-Fidelity Exploration

Mid-Fidelity Standardization

High-Fidelity Polished Version

Design System & Documentation

Component Library & Developer Documentation

Key Learnings

What I Learned

AI Will Fail—Design for Failure

Collaboration Shapes Better Design

Personal Growth Through Initiative

Program Outcome

Delivered

Impact

Recognition

Live Demo

More From This Director

Monple

PM Accelerator

MBITS Innovations