I didn’t start with a Figma file. I started with a bit of frustration. The early agency sketches came in. I didn’t like them. They looked like every other RO in the market. The explorations had curved fascia, top-mounted tap, hollow bezels, and shiny chrome. It was ornamental design pretending to be premium. Not the kind of product I’d want in my kitchen.
So I started sketching myself. I didn’t have formal industrial design training, but I’d done enough mechanical drawing in engineering college to start. And I spent nights watching Udemy and Skillshare tutorials on industrial sketching. Just enough to be able to speak the language of form, contour, and proportion with manufacturers and agencies.
The goal was simple: Make a machine that looked modern but wouldn’t age poorly. Think Audi, not Toyota. I avoided organic shapes, rounded faces, and over-designed surfaces. No visual noise. No floating panels. No shapes that existed just to look different.
What we did instead:
Rectangular body with softened edges
Clean vertical symmetry
Flush buttons
Tray that felt integrated, not an afterthought
It needed to be a silhouette you could remember. Something distinct enough that even without a logo, people could recognise it. We didn’t want people to say, "Nice RO." We wanted them to ask, "What brand is this?"
Most water purifiers in India look like they were made by 20 different product managers in 2 different years. You can’t tell one brand from another. So we made silhouette our brand.
The key shape: a vertically stretched rounded rectangle that was balanced, symmetric, and planted.
I looked at Braun for minimalism, Apple’s Macbook line-up for surface finish and visual restraint. Polestar (my favourite) for geometry and architecture from Copenhagen for calm and thoughtful facades.
We avoided any chrome, gloss, or backlit icons. We also rejected traditional RO shape conventions: bulges, transparent covers, and stickers with feature callouts.
Collaboration with design agency & manufacturer
We worked with a design agency based in Pune. They helped us convert our sketches and moodboards into CAD-ready forms. Initially, there were 4–5 design directions. We discarded all but one.
I didn’t want a generic product that got lost in a showroom. I wanted a machine that looked like a home appliance but behaved like a piece of furniture. Something that felt placed, not installed.
We used foam-board models to iterate form and we collaborated with the manufacturer for 3D printed prototypes to check fitment, assembly and final look.
During this phase, we also started factoring in DFM constraints: injection moulding draft angles, parting lines, and snap-fits. More on that shortly.
We made 10+ foam board prototypes. At first, it looked bulky. The filters we used were larger than standard OEMs. But then something clicked. Most kitchen light comes from above. Shadows fall behind the machine.
So we heavily curved the rear surface. The shadow created an illusion of lightness. The machine now felt 2 inches shallower than its actual depth.
Later prototypes included 2 FDM (3D printed) shells for touch testing, internals held with temporary fixtures, and real-world counter placement checks.
Of course, it wasn’t rosy picture either. In the early prototypes button feedback was inconsistent, the tray was too shallow and even the LED ring light wasn’t glowing uniformly leaving blind spots. We re-tooled the button domes, re-designed the tray to be vented and added optical light guides with diffusion for uniform glow. Amongst 50+ other fixes.
We started Design for Manufacturing (DFM) while finalising the CAD. Our manufacturing partner in Pune helped identify potential issues early. We changed the front fascia to a snap-fit to avoid visible screws, redesigned the tap cavity to allow proper tool clearance, and tweaked the button structure to reduce wobble.
For touch to work, the PCB and LEDs had to be mounted on the front fascia, which required a wiring harness to run across the body. To simplify servicing, we made the front panel removable and mounted the PCB on a front panel. This added complexity, but the product felt cleaner and more integrated as a result.
CMF decisions
Before we talk about what we did, let’s talk about who got it right.
Braun’s appliances that didn’t demand attention. The finishes were selected so the user didn’t need to think about them. They just worked. Matte textures, soft curves, no unnecessary gloss. Their surfaces were for daily use.
Dyson made surface finish a signature. You could see the airflow. See the dust. Their materials built confidence. The gloss had a purpose. The click of their attachments, the feel of their buttons.
Xiaomi took minimalism and scaled it. Their air purifiers, water dispensers, even smart lamps shared a disciplined CMF vocabulary. Just clean materials, in neutral colours, with low-shine finishes. That’s what made them feel premium without pretending to be expensive.
That’s what we tried to do at Native.
We picked every surface based on where it would live: Indian kitchens. Not Pinterest kitchens. Real kitchens with oil splatter, steel utensils, damp cloths, turmeric in the air, and zero time to pamper appliances.
We chose matte ABS with a satin texture for the body. Gloss might look good in marketing photos, but in real kitchens, it reflects harsh tube lights and scratches within days. Satin stays balanced, feels clean, and handles rough use better. For the tank, we used food-grade PP. The tap lever is polycarbonate for stiffness.
We launched in black. I was inspired by the Space Grey MacBook. It hid dust. It absorbed fingerprints. And it didn’t change colour under tube light or yellow with age.
Even the logo was debated. We made it small. Just big enough to know it’s there. But not loud. We wanted the machine to be the hero. Not the badge.
Remember what I covered in part 1 of this series? Our stance with design was simply to not try hard. We kept our language with strong silhouette, minimal face, soft contouring on body edges, and LED glow ring as visual anchor.
This was the shape that defined Native.
From first sketch to shipping product, we spent months iterating on form, materials, usability, serviceability.
We launched Native M1 and M2 in October 2023. We rolled out on our app first. Then Amazon. It was surreal to see the product on shelves. We were proud and overwhelmed and nervous and excited and speechless. It was honestly a 100 emotions coming together for all of us.
The moment that hit home for me was that my parents could finally tell their friends what I built. Not an app. Not a website. But something they could walk up to, touch, and use.
In the next part, I’ll share what it meant to build the team. And what I learned transitioning from digital to physical.
See you there.
