02 Business Requirements
Why is this required?
Avoiding the "Black Box". BOSCH was preparing to launch a complex fleet management service involving thousands of IoT devices, video telemetry, and cloud infrastructure. Launching such a system without visibility is a massive commercial risk.
The strategic mandate was not to fix an existing problem, but to prevent one. We needed to build an "Observable System" from Day 1 to ensure the Go-To-Market (GTM) strategy could succeed.
🐣
The Trust Gap
As a new entrant, BOSCH needed to prove reliability to early clients immediately. We couldn't afford "billing disputes" later; we needed a system that justified every dollar of service before the first invoice was sent.
🧑🏽🏭
Operational Scalability
The DevOps team needed to be ready to handle thousands of devices instantly. They required a tool that would prevent "alert fatigue" before it even started, allowing them to diagnose issues in a system that didn't exist yet
03 Research
Navigating Ambiguity (Zero-to-One)
Since this was a new product, there were no "users" to interview and no existing errors to analyze. The challenge was purely Ambiguity. I had to design for data that hadn't been generated yet.
We facilitated three rounds of intensive offline workshops (One was already done before my joining) with stakeholders from DevOps, GTM, and Logistics.
Round 1
Broad Pillars: We categorized vague complaints into three pillars: System Health, Revenue Credibility, and Operational Scale.
Round 2
Metric Drilling: We drilled down into specifics. For example, "Is the device working?" was broken down into distinct metrics: Trip Connectivity (Boolean), GPS Accuracy (Range), and Video Upload Success (Rate)
Round 3
The Ownership Model: We mapped every single metric to a business outcome either Cost Cutting (DevOps efficiency) or Revenue Generation (Billing justification). This created the "KPI Ownership Model" that became the backbone of my Information Architecture.
Since this was a new product with an aggressive 8-week launch timeline, we couldn't afford a traditional linear process (Research → Design → Handoff). Instead, I was put in the Arena, a pilot testing zone where Hardware, Firmware, Logistics, and Cloud teams sat together.
The process was not "hard and fast research". It was iterative, parallel, and high-velocity.
To keep up with the engineering pace, I leveraged Figma Make to generate high-fidelity prototypes instantly. This allowed me to bypass slow wireframing cycles and test complex interaction flows with DevOps engineers in real-time.
Design didn't happen in a silo. I would prototype a visualization/module (e.g., a latency heatmap), and then immediately sit with the Cloud team to fetch queries from the database. We tested live: Can we actually derive this KPI? Is the query too heavy?
Managing Trade-offs: This "fun, super-rapid handoff" style meant we were constantly negotiating trade-offs between "Ideal UX" and "Available Data," ensuring that every screen I designed was technically feasible from Day 1.
04 Design Goals
Need for Operational Efficiency
North star Goals
Give engineers "Context," not just "Alerts." The system must answer why a failure occurred (Root Cause Analysis) without manual digging.
Connect the physical reality (Installation) with the digital outcome (Billing) in a single view.
The GTM dashboard must function as a "Proof of Value" tool, visualizing reliability to secure early contracts.
05 navira Overview
One hero system to do it all
06 Interface Design
Function is important, so is the form
Every single UI element had a design decision
Conversational onboarding
The platform uses a four-question conversational flow to define the product idea, sensory preferences, dietary constraints, and desired budget/profit margin. Kai takes care of rest of the things!
Kai
AI Agent of Navira
BASELINE FORMULATION
A Visual Lab for Creative Experimentation!
Reactive Guidance
Experimentation as a Conversation
ASK KAI
Experimentation as a Conversation. Every experiment is tracked and secured by automatically creating new versions (e.g., gfc_v1), preserving previous attempts for later comparison
IMPACT SIMULATOR
The Impact Simulator (a radar chart) provides immediate visual feedback on the trade-offs resulting from any substitution across five key parameters: cost, nutrition, sensory profile, shelf life, and compliance
Quality Degradation Simulation
A critical feature allows users to simulate how product quality degrades from sourcing till expiry. The system tracks physical, chemical, and microbial properties based on customizable storage conditions (like temperature and humidity)
Navira also performs Digital Final Product Analysis against official compliance standards (like FSSAI thresholds) for key parameters such as protein, moisture, fat, and microbial counts
market intelligence
Connecting Your Lab to the Market. Kai generates sales projections, visually positions the product against competitors based on price and sales volume.
LOOPING CX INSIGHTS
Provides proactive analysis of consumer data, including peak consumption times and flavor preferences for the target demographic
sUPPLIERS
Kai aids the user by automatically listing potential suppliers, providing access to their Certificate of Quality
INVENTORIES
An intuitive, visual overview of your/rental storage, showing available space and highlighting items expiring soon to help reduce waste and manage stock effectively
Kai proactively checks inventory and storage capacity (e.g., cold storage unit availability) for planned manufacturing dates
NOTEBOOK
Designing for Relationships, Not Just Recipes.
Serving as the "single source of truth," the notebook tracks all project parameters, including financial metrics, launch dates, and strategic objectives. It allows founders to add teammates (sales, marketing) who can contribute comments and notes to ensure organizational alignment
08 Impact
Navira Cooked!
These metrics are projections derived from research insights and the founders' perspectives on the MVP, showcasing its potential impact.
15x
Faster than 1 physical trial. Meaning, founders can perform 15 digital experiments for every one physical trial.
4-6 weeks
90% Reduction
Time to Market
85%
Reduction in R&D Cost
eliminating most physical trials and external lab fees
09 Challenges
Challenges make design fun
ITS SCIENCE 🤯
Day 1, I struggled in the Lab. If founders know 2 out of 10, I was below zero. Today, I cleared most ambiguity and worked with R&D Experts on a formulation (patent releasing soon). Isn't that the best part of UX?
Intuitive
B2B UX is a different world, and one of the main goals was no training and onboarding time.
BUILD FROM SCRATCH
Twin all the physical process to the digital ones while simultaneously designing a futuristic North Star vision
being Whimsical
Product must demonstrate better experience, innovative, something fun, something creative, focusing too heavily on pure engineering functionality makes one less of a designer
Shipping Speed
The entire project, including the initial MVP build, external reviews, critical pivot, and final prototype design, was executed within a tight timeframe, meaning I hardly got 35 days for finishing all my work
10 Learnings
Every challenge was a learning
It’s been an incredible journey working on this project and collaborating with so many R&D Experts and founders at various scale. One big mistake which I did is doing it alone. This is my first solo project on such a big scale (subjective). Given its scale and impact, I learned a lot throughout the process. Every challenge faced and solved was a valuable lesson.
Thanks for reading!
More to build, more to share
