Intro

The most frequent question we get asked by leaders from the financial services sector is:

“How can we integrate AI into our processes and increase ROI?”

This article outlines our approach.

The Problem

A financial services firm approached us last year to help improve their loan underwriting process.

Their existing approach involved analyzing information provided by each loan applicant along with their credit history to assess whether or not that applicant met lending criteria.

Even when facilitated by digital processes and “rules-based lending models”, the underwriting process was largely a manual one, often took multiple days to complete, and involved many people.

The Challenge

How can we develop a credit risk model that helps underwriters determine whether to lend to a specific applicant… in seconds.

The Solution

We developed an algorithm powered by machine learning and AI that deeply analyzes millions of historical loan applications and loan performance data, recognizing patterns that can predict how a new loan applicant is likely to perform in the future, all within seconds.

The ROI

This new model is forecasted to reduce current default rates by 5 to 10%, improve profits, reduce portfolio risk, and shorten the underwriting process from days to seconds.

Our 4-day Approach

We use Google Design Sprints to research, design, prototype and test a new solution to transform businesses using artificial intelligence.

With this approach, ideas can be refined and evaluated well in advance of making any technology investment. This increases effectiveness and dramatically reduces risk.

Post-Sprint

After the sprint, the team will have a clear understanding of whether the solution prototyped is:

Viable from a business perspective

Feasible technically

Desirable to the end user

If the outcome is positive and Senior Management gives the go-ahead, we work alongside the organization’s teams to iterate, develop and deploy the solution into production.

We also provide post-launch support, documentation, and user training.

Interested in running a Sprint?

Please reach out to us to learn more.

With gratitude,

Jameel & Monish

Jameel Somji, CFA is Founder & CEO of SpiNovation Labs, a product design lab that applies 4-day Google Design Sprints, a proven methodology for solving problems through designing, prototyping, and testing ideas with users rapidly with as little investment as possible in as real an environment as possible.

Monish Gandhi is Founder & CEO of Gradient-Ascent AI, a company that helps companies get started with applied machine learning & AI.

This is a story about a Design Sprint. If you’ve never heard of that, you can learn more here or read the book.

Intro

In May 2018, a real estate developer approached me to determine if they should build a $30 million townhome community for active seniors aged 59–74 in Niagara Falls, Canada.

Why?

A year prior, the developer built a retirement residence in a different city but only managed to achieve a 50% occupancy rate.

Why?

A desk research report provided by their bank’s financing partner recommended that 80% of the units constructed be one-bedroom and the remainder, be two-bedrooms.

After constructing and opening the residence, the developer realized that there was significantly more market demand for two-bedroom units.

To avoid making the same mistake, the developer wanted to apply a more rigorous research approach, to inform the real estate development strategy for their new project, prior to breaking ground.

I suggested running a Design Sprint.

Sprint Challenge

Prototype & test a new townhome community concept with seniors in 5 days.

Sprint Participants

• CEO
• Architect
• Head of Marketing & Sales
• Head of Operations
• Designer
• Sprint Master (Jameel)

We also consulted with these experts during the sprint:

• Two local real estate agents, a seniors moving company, a local senior home, a commercial banker, a local senior recreation centre and an urban planner at Niagara Fall’s city hall.

The CEO Wanted these Questions Answered

1. Are seniors interested in renting a unit in the townhome community?
2. Is the rental price affordable?
3. Is there demand for 1, 2 or 3 bedroom units?
4. Is the location desirable?
5. Is the design, layout and finishings of the units desirable?

How we built & tested the prototype in 5 days

After mapping out the challenge and talking with experts on day 1, each participant sketched and presented their prototype concepts for the townhome development on day 2.

On day 3, the team voted on the winning concept and created a storyboard of the design and features.

On day 4, we built the prototype of the townhome development in one day.

How we Found Seniors to Test With

We asked a local senior recreation centre and bingo hall to post this ad:

We received 10 responses in 3 days and invited 5 seniors that fit our target profile to give us their feedback on the prototype on Day 5.

Customer Feedback

All 5 seniors said they would NOT rent a unit in the townhome community because:

-They didn’t want to climb 4 steps to access the unit

Seniors looking to downsize don’t want to move into a place that has stairs, to prevent the risk of falling in the future.

-The townhome didn’t have underground parking

Winters in Niagara Falls are severe. Seniors wanted their vehicles protected from the extreme cold, snow and ice.

Apart from the issues above, Seniors were happy with the pricing, location, layout, and amenities and preferred the two-bedroom unit.

Sprint Outcome

The seniors’ townhome community concept was an efficient failure.

The 4-day Iteration Sprint

The CEO didn’t want to amend the design of the development because it would materially impact the project ROI, so we decided to test the townhome concept with a younger demographic.

Customer Feedback

All five customers said the rental price was too high for a townhome

NOTE: Customers mentioned they would rather rent a larger detached multi-story home in the same area for a just a bit more money.

Iteration Sprint Outcome

The townhome community concept was an efficient failure because the rental price was too high.

Business Model Pivot

Given the customer feedback, the CEO decided to pivot from offering units for rent to selling condos. See the text message below from the CEO.

Conclusion

The sprint helped the CEO confidently pivot from offering units for rent to offering condos for sale to investors who were priced out of the Toronto real estate market and were looking for a more affordable place to invest.

NOTE: Townhome prices in Toronto go for $1–2 million while the asking price in Niagara Falls was $0.2–0.5 million.

All units in the townhome condo project sold out within six weeks of opening pre-construction sales.

The project completed construction and the owners took possession of the units in October 2022.

Jameel Somji, CFA is the Founder & CEO of SpiNovation Labs, a product design lab that applies a proven methodology for solving problems through designing, prototyping, and testing ideas with customers rapidly to determine if they are worth investing in.

Would Seniors Rent units in a $30 Million Senior Townhome Community Development in Niagara Falls? was originally published in Sprint Stories on Medium, where people are continuing the conversation by highlighting and responding to this story.

Intro

In 2016, I helped 13 companies from McMaster University’s Startup Incubator (The Forge), prototype and test 13 new product ideas with real customers over the course of 15 days.

One startup I worked with was founded by a PhD student who discovered a way to analyze medical images automatically with the help of AI.

Sprint Goal

Prototype an AI powered software that is trained to segment medical images in seconds.

Sprint Participants

• Founder & CEO
• Medical Researcher
• CTO
• Designer
• COO
• Sprint Master Jameel

Note: We also talked to experts about accounting and product marketing unknowns. We also talked with an AI expert on how to acquire and structure the needed data to train the AI.

The team wanted to find answers to the following questions by the end of the sprint

1. What types of researchers would be interested in the tool?
2. How much data is needed to train the AI?
3. How should the data be structured?
4. How much are researchers willing to pay for the tool?

How we built the prototype

After talking with experts, each team member individually sketched and presented their prototype solutions.

We then voted on the best solutions and created a storyboard for the prototype below:

After designing the storyboard, we built the prototype in one day using Photoshop & InVision:

What we learned from customers

• Customers require a way to view and refine the segmented medical images after being processed by the software
• Brain researchers are not a good fit for this product because alternative software exists.
• Don’t target researchers who are engineers – they prefer to segment images themselves as they are technically trained to do so.

“The Sprint process forced us to build a prototype and learn from customers. We would not have gotten this far on our own”.

-CEO, NeuralSeg

Interested in Running a Sprint?

Please reach out to me to learn more.

Jameel Somji, CFA is Founder & CEO of SpiNovation Labs, a product design lab that applies 4-day Google Design Sprints, a proven methodology for solving problems through designing, prototyping, and testing ideas with users rapidly with as little investment as possible in as real an environment as possible.

Intro

Around Christmas last year, I was introduced to the CEO of a music blockchain startup. We had something in common….we were both Deejays, and also CEO’s of companies we founded, her seven times over!

She had just come back from Amsterdam. Her company was sponsoring a big music festival there to spread the word on how her company can help independent artists get paid more by leveraging their fanbase.

FYI…Artists don’t make much money off streaming platforms if they are not the likes of Drake, Rihanna or Ed Sheeran.

Sprint Challenge

Prototype a platform that enables artists to compensate fans for promoting their music (using the Etherium Blockchain).

Sprint participants:

• CEO
• CTO & Blockchain Developer
• Information Architect
• Head of Artist Relations
• Head of Marketing & Sales
• Visual & Interaction Designer
• Sprint Master Jameel

*Note: Experts were also brought in to shed light on finance, legal and customer unknowns. We were also fortunate to talk with a co-founder of Ethereum on how best to develop the platform.

The team wanted to find answers to the following questions:

1. Are artists willing to buy the company’s crypto token to transact on the platform?
2. What is the best way for fans to promote artists?
3. How much would artists be willing to pay their fans to promote them?
4. How to structure Ethereum smart contracts between artists & fans?
5. What are the technology risks associated with the platform

How we built the prototype

After talking with experts, each team member individually sketched and presented their prototype solutions of the platform.

Once a solution was voted and settled upon, we built the prototype of the website landing page using Google slides.

What we learned from customers

• Independent Artists like the product idea but don’t think it will work.
• Artists are reluctant to leave popular streaming platforms for something new.
• Artists have no qualms about buying and transacting in Cryptocurrency.
• Independent artists won’t pay a subscription fee for the platform. They are more interested in a revenue sharing model.

“We see great value in the Sprint process and are going to run one again on our own, very soon!”

-CEO, Music Blockchain Startup

Interested in Running a Sprint?

Please reach out to me to learn more.

Jameel Somji, CFA is Founder & CEO of SpiNovation Labs, a product design lab that applies 5-day Google Design Sprints, a proven methodology for solving problems through designing, prototyping, and testing ideas with users rapidly with as little investment as possible in as real an environment as possible.