B2B SAAS Dashbord (Part II)
MyFleet is a story of observing real-world customer behaviour to design the right interventions at the right time. By spending time on the ground with EV fleet operators, we uncovered gaps in how they tracked vehicles and managed drivers. Within a week of it's launch, 75% of users adopted it—showing how thoughtful design can drive real impact.
What does Alt do?
We help fleets lease commercial electric vehicles (EVs) across 20+ Indian cities. With ₹200 cr worth of vehicles to manage, real-time monitoring is non-negotiable to keep them running smoothly.
Seen Zepto, Zomato EV drivers riding flashy two-wheelers? Or that three-wheeler cargo pickup waiting outside your house with a delivery package? Remember the last time you booked a vehicle via Porter to pick up your stuff? Yes, I’m talking about those vehicles.
Problem Statement
Managing a fleet was a manual and inefficient process. Fleet operators struggled to track vehicles in real time, prevent misuse, and manage driver accountability. Scattered workflow ensured there was no single source of truth.
Lack of visibility of routes: there was no way to track the routes drivers took; drivers often took detours, took up side gigs, and used the vehicle to run personal errands.
No insight into driver behaviour: there was no visibility into driving behaviour; drivers would rashly accelerate, overspeed, jump traffic lights accumulate unnecessary challans.
Geofence violations & risk of vehicle theft: there was no way to ascertain if the vehicles were operating in the designated neighbourhood/ geo-zones.
In the rare case of vehicle theft, an on-ground manual recovery team had to be deployed. This manual process would cause delay in response time and increase the risk of vehicle theft.
These inefficiencies eroded trust between fleet operators, supervisors, and drivers, making safe and timely deliveries difficult.
Objective
Our goal was to help fleet operators make faster, data-driven decisions by giving them real-time visibility into their EV operations.
We set out to redesign the MyFleet dashboard to streamline manual workflows, reduce dependency on calls and WhatsApp updates.
My Role
👋🏼 Hi! I’ve been a product designer at Alt for over a year, leading my first project from discovery to delivery.
To understand our users’ challenges, we:
Conducted 8 user interviews
Shadowed fleet operators
Visited 12+ vehicle dealerships
Observed parking hubs
Objective
Our goal was to design tools that could help fleet operators:
Improve vehicle uptime by spotting vehicle issues early.
Give better visibility into driver behaviour, reward good driving, and act fast when things went wrong.
My Role
👋🏼 Hi again! This was my second phase of MyFleet project, where I led design for a set of safety & performance tools:
Design and test custom alerts & driver scoring models
Prototype early immobilisation flows for vehicle theft or risk
Research Findings
When we investigated low customer satisfaction with the legacy dashboard, we found that while it offered a general performance overview, it failed to surface critical alerts and real-time trip data.
Fleet operators across India—managing 200–700 EVs daily—were relying on manual calls, google sheets, and WhatsApp groups to stay updated. These methods were error-prone, and time-consuming.
—Fleet Founder, Delhi
In over 8 interviews with fleet owners, city managers, and ground teams, we uncovered recurring pain points:
Vehicles often left designated zones without alerts.
No reliable way to confirm completed trips.
Side gigs and personal detours went undetected.
No ability to set geofence & monitor breaches.
These findings pointed to a larger insight: operators needed real-time visibility and control, not just data summaries.
User Stories
We condensed the above problems and insights from our research into simple user stories:
As a fleet operator, I want to see live vehicle location, so that I don’t have to call drivers to track their trips.
As a fleet operator, I want to set geofences, so that I get notified when a vehicle leaves its zone.
As a fleet manager, I want to receive real-time alerts for vehicle misuse or delays so that I can intervene before it affects deliveries.
A typical parking hub of our Fleet Operator, consisting of 20–30 vehicles
Solution (Design) — Introducing MyFleet!
Myfleet is our subscription based fleet-monitoring dashboard given exclusively to our customers or B2B operators along with their lease.
Existing Screens
Here are our existing screens. Our exising dashbaofrd "FleetOS" was made as an internal tool, which we later gave access to our customer as the requests increased. sooon enough we realised that it could not hold the complexity and granulaity of the peoblrms fleetops faced
Add it in the UI also. Some things it lacked:
vjnvjfv
jrnvjrv
vjfnv
A typical parking hub of our Fleet Operator, consisting of 20–30 vehicles
Introducing Trips & Stops
PROBLEM
#1
Lack of visibility of routes: there was no way to track the routes drivers took; drivers often took detours, took up side gigs, and used the vehicle to run personal errands.
To solve the visibility of the route drivers took, we looked beyond live location and started storing the coordinates to form a trip. We looked at various different ways we could now visually show where all the vehicle travelled.
One insight we observed while shaowfing hub operator was that driver would oftent ake detour to run personal errand sor take up side gigs during work hours.
Knowing that the fleet operator would constantly guess the time taken to complete trip, where the driver stopped for a break, the duration of the break allowed us to do the heavy lifting by categorising trips and stops!
Trips
Certain vehicles would run more than the expected delivery distance. This would happen because a lot of times the drivers go off route to do their personal work. Trips would catch these instances.
Stops
We classified static readings as “Stoppages” which could indicate:
Driver waiting at a traffic light
Loading/unloading goods
Lunch/water break
Geofencing
PROBLEM
#2
Geofence violations & risk of vehicle theft: there was no way to ascertain if the vehicles were operating in the designated neighbourhood/ geo-zones.
We designed a comprehensive detailed analytics dashboard on geofence breaches. Showing these breaches helped in bringing visibility.
The UI needed to inspire confidence by showing all the breaches made to the geofence so they could call the driver & alert him, when needed.
To validate our assumptions, we travelled to various hubs in Delhi to test an early prototype. To our immediate surprise, none of the participants faced friction in setting up geofences, despite our expectations otherwise. The feature resonated well with participants, confirming our intuition around designing for speed.
Over-designing Filters
PROBLEM
#3
Lack of customisable filters: filters were too basic to use & hidden away making it impossible to quickly find relevant vehicles, trips, zones, or violations.
The goal of the filters was to slice and dice the cluster of data. I assumed that more granularity and customization would reduce the operator’s uncertainty. However, after testing many iterations of this flow, we repeatedly observed that:
Operators only wanted high-level information such as city and vehicle category to see the number of vehicles in a bucket (e.g., showing 27 results).
Operators weren’t too worried about granular-level data such as the number of vehicles with battery <25% in operating in Okhla #14 geofence.
These observations made us realise that too many options overwhelmed them. After a third round of poor testing, I scrapped the idea in favour of a high-level quick filter.
Map, charging and service garage
Users can view all vehicles on a map with current locations and status indicators.
They can filter the vehicle list based on various criteria (OEM, type, city, status, mobility)
Access a summary of active alerts categorised in the cards (e.g., service due, low battery).
Usability Testing
write stuff
see samrat's comment,
hotjar report
tested immobilisation prototype (framer) on ground
iterations (if any else case study 1 mein filter wale pe kar sakte ho)
Impact on Business & Users
Before launching this set of update on MyFleet, fleet operators struggled to keep up with one zone in a day, manually tracking locations, coordinating with drivers, and picking endless calls. It was slow, inefficient, and frustrating. With MyFleet, our B2B customers now manage their entire fleet vehicles across zones without daily hassle.
By fixing their biggest roadblocks, we made fleet operations smoother, faster, and scalable. And as they expanded, they leased more vehicles from us, fuelling our own growth in return. Their success became ours, proving that when we solve the right problems, everyone wins.
Key Performance Indicators
78% reduction in manual coordination and tracking methods.
75% of our Fleet Operators use Myfleet on a daily basis.
Team :)
Taking sole ownership of the project was the most riveting experience of my life. I learned so much about how the EV ecosystem functions, how different teams operate, and about the on-ground reality of fleet management and its share of nail-biting stress. I’m happy we could make our customer’s job just 1% easier.
I couldn’t have done this without the help of my collaborative high-agency team. A big thank you to the incredibly talented people for supporting me every step of the way.
Rohit Kashyap (UX Design & Research Intern)
Yashoheet Sethi (Product Manager)
Prateek Jha (Full-stack developer)
Sharad & the entire Data Science team.
+founders, engineers, operations team. And to our customer who took out time to sit with us and graciously offered us to shadow them :)