Back

Manny AI

2024

Overview

Manny AI set out to transform how garment factories plan and execute production. Typically, production managers depended on spreadsheets and disconnected systems to manage highly complex, multi-stage manufacturing workflows.

While a basic digital prototype existed, it had no real UX or UI structure. It wasn’t scalable, intuitive, or usable in the realities of live factory operations.

I was the sole Product Designer, working alongside developers and client stakeholders to reimagine Manny AI from the ground up.

The goal was clear: create a single system where planners could forecast production capacity, machinists could execute tasks with clarity, and the AI could predict and adapt to real-world bottlenecks as they emerged.

Contributions
Design Lead — Interaction Design, Visual Design, Prototyping, User Flows, Copywriting

Timeline
3 months

Tools used
Figma, Adobe CC, Miro

HOW MIGHT WE

How might we transform garment manufacturing’s dynamic, multi-stage production workflows into clear and actionable systems, ensuring both planners and machinists see what matters, when it matters?

Discovery & direction

Before defining solutions, I worked with the team to map out the existing production workflows and pain points inside a Miro board. This allowed us to see how scheduling, communication, and manufacturing stages broke down across the factory.

Garment manufacturing has unique needs that general project management software can’t handle. A single order isn’t just a task, it’s a series of dependent stages, where delays cascade and resource management is a constant balancing act.

Planners needed to understand not just what was happening now, but what might happen next. They needed to simulate new orders, model workforce implications, and spot risks before they became problems.

Initial workflow mapping and challenge discovery.

IMAGE

Core objectives identified

Operational forecasting

Give planners a predictive view of factory capacity across orders, workforce and timelines.

Real-time execution feedback

Build a feedback loop from machinists to planners, allowing live adjustments to schedules.

AI-driven task prioritisation

Empower machinists with clear next steps generated by the system based on skillsets and production needs.

System scalability

Design workflows and interfaces that could scale naturally as factories and production volumes grew.

System feedback loop — planning, prediction, execution.

IMAGE

Approach & execution

Designing Manny AI required building an interconnected planning system from the ground up.

Rather than treating each feature as a standalone tool, I approached the Sell screen, staff scheduling, product setup, and task management as an integrated operational ecosystem. This would ensure that every planner action could ripple intelligently through the entire production workflow.

SELL SCREEN

The first major design move was the Sell screen, a tool that allowed planners to model potential orders against live factory capacity.

Adding an order would immediately show its impact on current schedules, helping managers make confident commitments to clients.

STAFF SCHEDULING

Alongside the Sell screen, I redesigned the staff scheduling experience, linking workforce allocation directly to live production needs.

Planners could now visualise how staff shifts and machine assignments evolved dynamically with changing order volumes.

Discovery & direction

Before defining solutions, I worked with the team to map out the existing production workflows and pain points inside a Miro board — capturing how scheduling, communication, and manufacturing stages broke down across the factory.

Garment manufacturing has unique needs that general project management software can’t handle. A single order isn’t just a task — it’s a series of dependent stages, where delays cascade and resource management is a constant balancing act.

Sell screen: real-time factory capacity simulation.

IMAGE

Visual design

PRODUCT SETUP AND TEMPLATES

Product setup also needed rethinking.

I introduced a new workflow where planners could build multi-stage manufacturing processes, defining deliverables while visualising timeline impacts at every stage.

Templates were created via a Product Directory, allowing fast reuse of complex workflows.

TASK BOARD

For ongoing production management, I redesigned the Task Board to bring kanban-style clarity to live manufacturing stages.

Instead of static spreadsheets, planners could now see work moving, spot delays visually, and act immediately.

Two worlds connected — planner-side control and machinist-side execution.

IMAGE

Companion app for machinists

While the planner interface gave planners a clear view of schedules, day-to-day execution still relied on machinists having fast, reliable task visibility.

I designed a companion app to make task management on the floor simpler, clearer, and more responsive.

Task selection and tracking

Machinists could follow a recommended task, generated by the AI based on their skills and production needs, or select from a full task list, organised by status: Not Started, In Progress, or Completed.

Each task tracked live progress with a built-in timer, giving planners real-time data on how production was unfolding without the need for manual updates.

Real-time production planning

By linking task execution directly to production planning, the app helped planners adjust schedules dynamically, spot delays early, and keep the factory moving without disruption.

Deviation between expected and actual task times surfaced automatically, allowing planners to predict bottlenecks before they escalated into wider production issues.

Machinist task selection, live progress and timing.

VIDEO LOOP

Visual system & branding

The visual design needed to feel calm, modular, and production-ready.

I built a component-driven visual system focused on clarity, glanceability, and repeatable UI patterns.

As part of evolving the Manny AI brand, I also redesigned the logo. Clusters of connected nodes forming the abstract letters “M,” “A,” and “I” which represented dynamic production flows and the intelligence driving the platform.

Manny AI visual identity — modular and intelligent.

IMAGE

Build & implementation

Manny AI was already in use by factories while I was designing, so implementation happened in parallel, with new features introduced incrementally into the live platform.

I ran regular workshops with the client and development team to align design intent with technical feasibility, often adapting workflows on the fly based on current capacity and constraints.

We built using DaisyUI’s component system, working closely in live design–dev sessions to extend patterns where needed. For example, co-designing a custom timer overlay when standard modals couldn’t support live task tracking.

This collaborative, iterative process kept the system coherent and shippable, even as the product evolved in real time.

Outcomes

Increased scheduling efficiency
In early user testing, planners reported a 40% reduction in scheduling time.

Real-time visibility from the factory floor
Machinists were able to onboard without formal training, completing first-shift workflows successfully.

Strengthened client confidence
Planners began using the Sell screen in live client negotiations, boosting confidence when taking on new orders.

IMPACT

Manny AI gave the factory a predictive operational system for the first time, not just tracking what happened, but anticipating what would happen next.

KEY LEARNING

Great operational systems remove uncertainty. Manny AI wasn’t about managing complexity. It was about creating clarity, predicting action exactly when it matters most.

HOW MIGHT WE

How might we transform garment manufacturing’s dynamic, multi-stage production workflows into clear, actionable systems—ensuring both planners and machinists see what matters, when it matters?

Project Takeaways

Predictive systems, not passive tools

Operational products need to help users see forward, not just look backward.

Clarity over choice

Clear task prioritisation for machinists drove stronger floor execution without adding complexity.

Trust through simplicity

The most valuable feature wasn’t more functionality. It was making complexity navigable, understandable, and actionable.