# Wint.Ai


# My role
To research a solution and provide a design.

### General Overview
Wint Wi has been a long-time client of mine since 2014. In recent years, they made notable growth and partnered with several companies among some of the leading techs in Israel (Bank Hapoalim, Google, Wix, HP, and many more).

Wint recently recruited a sales team in the United States and launched new test sites in Eastern Europe with HP, while they keep working on their 3rd generation product.


**WintWi Labs: Rosh Ha'ayin, Israel.**

## Data and potential outcome
Wint algorithm analyzes and translates the water flow rate into actual usage by understanding consumption patterns throughout time.
- From showers to Irrigation, Wint labels types of consumption.
- Identify people's daily pattern
- Alerts leaks in real-time via SMS
- Automate turns off valves during massive water loss and pipes explosion
- Can give insights about physical deterioration

And more potential features from raw datask

## Pain points
- Wint provides a minimal bootstrap hard to use dashboard
- The support team has massive calls from many clients requesting essential guidance through the web service
- The algorithm team require feedback to improve the AI
- Customers demand a custome and handy mobile controller/dashboard

## The 2nd generation device has a minor difficulty.
The device was designed as part mechanic-valve part IoT device remotely controlled using GSM Technolgy. As a result, the device has a few trade-offs that reflect the UI design process and make it challenging.

- To preserve battery life, Wint designed it to turn on a 'sleep mode' for 3-6 hours after big chunk of data during a critical leak
- Asleep mode turns off the communication circuit.
- Sleeping mode delays any command sent to the device.
- Sleeping mode relies on sets of rules while it's on.
- The valve can close only manually during the Asleep way.

## Human computer interaction
(Simple version)

# Typical WintWI customers:
- Home owner
- Gov company
- Corporate (+20 units)

# User research
To better understand how to solve real-time leak events and allocate human resources, I conduct research in three different types of client cases. My primary focus is on the corporate type and their facility manager's crew.

## First case: Wix

I meet Ido, seven years facility manager at Wix, A friendly, hard-working person who is in charge of the maintenance of Wix 17 buildings. Ido has five people on his team.

### On-site Interview findings
- Wix is located near the beach in Tel Aviv, spread over 17 small buildings across a 5-kilometer area.
- Most of their offices' infrastructure is rusty and old, causing leaks due to hummid conditions.
- The company has a 24/7 online support team.
- Ido has five people in his team; he sends a team member to check the system's critical alerts when he's not around.
- Ido tends to set the valve open even during a critical leak avoiding water interruption for the employees.
- Ido is aware of events on the Web platform but ignores the low leaks events.
- He has a trust issue with the system but relies on the critical leak feature avoiding valve closing remotely.

## Second case: Isracard
I meet Igal, a company veteran, working at Isracard for almost 20 years. He is a tech-advanced type of role who stays alert 24/7 and gets insight from the building systems.

Igal sends a team member to follow up with Wint's critical alerts during the weekend. Igal has a team of nine.

- Isracard has 22 units installed on 22 pipes for each floor.
- The building pipe system is ok, highly maintained, but far from new.
- Isracard has a busy restaurant and cafeteria facilities producing a lot of waste during the mornings and launch time
- The building has traditional work hours, from Monday to Thursday 09:00 am to 05:00 pm, the building is empty during the weekend.

## Third case: BSR


BSR known as a real estate investment group in Israel, mostly known for its five towers in the Ramat Gan area.

I meet Yakov, a two-year property manager in the group.
He is responsible for two buildings with 30 floors each, 1,400 sqm per floor divided to smaller offices in private ownership.

Yakov is continuously under pressure. His responsible for collecting management fees, different suppliers, and renovation processes in both buildings. Overall manages 60 Wint units.

Some of the issues that Yakov deals with are tenants that tend to have a luxury office and have big aquariums and fancy Jacuzzis.

Dealing with substantial water bills is something that Yakov is used to, but he wishes he could be able to track those exceptional cases to charge higher fees for using more water.

### On-site research key points
- The building, pipes, and infrastructure is brand new
- Yakov wants to get alerts and solve leaks asap with no delay
- Yakov wishes to compare day to day consumption in abroad overview to surveillance tentens
- Yakov wishes to assign a team member to a leak event.

# Conclusions
My thought significantly changed after the research.
I thought saving water would be an essential incentive for success, but I was wrong. Instead, the manager tends to listen to critical alerts avoids low leak alerts, and remote close.

The manager acquires more appreciation for the system through problems he solved (how many water taps he closed, toilet button he fixed, and pipes checked). The critical leak insights are how they justify Wint's coast and effectiveness.

> Visit https://wint.ai/

# Another problem

A general confusion is that the AI can detect multiple leaks in one unit. How is it possible when there is only one pipeline? - In commercial cases, the AI can stake up to unlimited alleged leaks, which confuses users.

# Solution

I focused on the UX rather than the UI, information is critical, and some customers prefer not to be controlled by AI, so they don't benefit from the 'Auto-shutdown' service.

My solution was a 'stacking mode,' where all leaks motorized without human intervention providing only user feedback.

In addition, I designed a full follow-up process for multiple team memebers.


# Deliverables
- Company presentation
- UI Design
- [Prototype](https://www.sketch.com/s/d524b413-f7a6-4153-934f-b75b17df1f55/v/m1Vdmg) | [Figma](https://www.figma.com/file/vSdfRgGCnDuBe5YMwFMYue/wint-19-8-18-wireframes?node-id=0%3A1)

## Final App
[WintWi App](https://apps.apple.com/us/app/wint-water-intelligence/id1475760554)


## The results (1-3mo)
- The Support team enjoyed a tremendous reduction in calls, focusing on customer success.
- Users enjoyed the control and complete transparency.

## Team Memebers
- Yaron Dycian CPO
- Itay Kidon, Full Stack Developer
- Matan Antebi, Software Engineer
- Maayan Raviv, Customer Success Specialist