AI Product Strategy · Assignment 3

One AI product to fund. The hard part is choosing the right problem.

A Bangladeshi funder will back a single serious AI product. Budget isn’t the constraint — judgment is. This is our case for wherethat product should aim: Bangladesh’s garment industry, and the real problem worth solving inside it.

Sector
Ready-Made Garment (RMG)
Discipline
Product Strategy
Focus
Problem Selection

The Case · 01

The mandate — and the decision hiding inside it

A Bangladeshi funder wants to back one serious AI/LLM product, built in Bangladesh, for a specific professional or high-value user group. With money off the table, the entire outcome rests on one choice.

So the work isn’t engineering — it’s deciding which problem is worth solving: which industry, which user, which workflow, and which risks matter most. The rest of this page is how we made that call.

The Case · 02

Why garments — bigger market, higher return

First decision: the arena. We weighed garments against agriculture and chose the larger, more strained market.

We chose Ready-Made Garment (RMG) over agriculture on market size and return potential. Garments are a ~$39 billionindustry — the world’s second-largest exporter after China, with $30B+ in annual exports.

Agriculture is smaller (~$12 billion) though growing ~9% annually. Bigger market, tighter margins, and a clear competitiveness crisis make garments the higher-ROI place to solve a real problem.

~$39B
RMG industry size
#2
Exporter globally, after China
$30B+
Annual garment exports
7,000+
Factories nationwide

The Case · 03

It’s deeper than defect detection

Our first instinct was manual quality control. Research showed that’s a symptom, not the root cause. Bangladesh is losing market share to Vietnam and China for structural reasons:

  • High operational and business costs
  • Persistent energy shortages disrupting production
  • Elevated bank lending rates
  • Infrastructure deficits versus competitors
  • Lower labor productivity, despite lower wages
  • Weak marketing and buyer communication
  • Political and economic instability
  • Difficulty demonstrating compliance to international buyers

The Case, In One Statement

Bangladesh garment factories are losing competitive ground to Vietnam and China due to high operational costs, infrastructure gaps, and lower labor productivity.

They also struggle to demonstrate labor, safety, and environmental compliance to international buyers — which directly costs them contracts. Without affordable, integrated tools to improve efficiency, reduce waste, and prove compliance, factories continue losing market share and export revenue.

The Approach · 04

Two questions for the Client — and what each unlocks

Before committing, we pressure-test the case with the funder. Each question is chosen to force a specific decision.

Beyond defect detection, what is the biggest cost or efficiency problem draining your factories' competitiveness against Vietnam and China?

Decision it unlocks

Whether quality control is the core problem, or whether deeper issues — energy cost, labor productivity, supply-chain delay, compliance burden — matter more. Determines which problem the product should actually target.

If we build an AI product to solve that problem, what would success look like, and how would you measure ROI or impact?

Decision it unlocks

The success metrics that matter to the funder — cost savings, faster turnaround, fewer rejections, regained market share, compliance speed. Determines how to design, price, and measure the product.

The Approach · 05

Mapping what we know, believe, and doubt

A CSD matrix — Certainties, Suppositions, and Doubts — turns the case into a plan. The doubts are the riskiest column: each one is an assumption to validate before a line of code is written.

Certainties

What we know

  • Bangladesh RMG industry is ~$39B; 7,000+ factories.
  • Factories rely primarily on manual quality control.
  • Vietnam and China are actively gaining global market share.
  • International buyers require documented labor and environmental compliance.
  • Compliance failures cost contracts and buyer relationships.
  • High operational costs and energy shortages are documented disadvantages.

Suppositions

Believed, not yet validated

  • An AI-assisted solution could meaningfully reduce defects and waste.
  • Compliance documentation can be partially or fully automated.
  • Factories will adopt new tech if ROI is clear within ~12–18 months.
  • Existing equipment (cameras, conveyors) can integrate with AI software.
  • A Bangladesh-built solution can be cheaper and easier to adopt.
  • Buyers would favor suppliers with AI-verified quality/compliance.

Doubts

Uncertain — needs validation

  • Which specific compliance issues matter most to which buyers?
  • What is the real willingness of factory owners to invest?
  • How much can AI realistically cut operational cost in practice?
  • Can one solution work across varied factory sizes and legacy equipment?
  • Would factories trust AI over human inspectors initially?
  • What price point is affordable to mid-market factories yet profitable?

The Opportunity · 06

Who we’re solving for

The buyer is the Factory Owner / Operations Manager — risk-averse, but decisive when the payback is clear. Everything below is stated in their terms.

Profile

Age
35–55
Role
Owner / senior ops or QA manager
Experience
10+ years in garment manufacturing
Location
Dhaka or Chittagong
Factory size
500–2,000 workers
Revenue
$5–50M annually
Exports to
Mainly EU and USA

Pain points

  • Losing contracts to Vietnam and China
  • Slow, costly manual inspection
  • Hard-to-prove compliance
  • Rising costs and compressing margins
  • Energy disruptions; slow production cycles

Goals

  • Cut production costs (15–25%)
  • Prove compliance easily
  • Win back contracts
  • Compete on price and speed
  • Reduce defects and waste

Needs

  • Affordable, integrated solution
  • Easy integration with existing equipment
  • Clear ROI within 6–12 months
  • Audit-ready compliance documentation
  • Staff training and local support

The Opportunity · 07

Where a Bangladesh-native solution wins

The incumbents are capable but ill-fitted to mid-market Bangladeshi factories. That misfit is the opening.

WiseEye (China)

Advanced fabric defect detection, widely deployed — but expensive and equipment-heavy.

Uster Technologies

Global textile QC leader; premium price, serves large factories.

Emerging AI startups

Mostly aimed at developed markets, not Bangladeshi factory realities.

The Opportunity

No locally built, affordably priced solution is tailored to Bangladeshi factory setups, equipment, language, and buyer-compliance needs — with local support and training.

The Result

A problem worth funding — defined, not assumed

Assignment 3 · Due 11 July 2026

  • A defined problem — cost, productivity, and compliance, not just defects.

  • A ranked set of doubts to validate before build begins.

  • One target persona the product is accountable to.

  • A clear market gap a Bangladesh-native product can own.

The focus stayed on the problem — the solution is the next conversation, and now it has firm ground to stand on.

Open the submission form