Ep:11 Ather Energy | Tarun Mehta on Engineering, Scale and India's Tech Future

ELI5/TLDR

Tarun Mehta, who co-founded the electric-scooter company Ather, spent 13 years building not just a product but a “machine that builds products” — and now he gets to use it. His big argument: India keeps thinking of “owning more of the supply chain” as owning factories and mines, when the real prize is owning the design — the chips, the software, the brains. He’s also banging a drum about standards: when India leads in some technology, it should write the rulebook the rest of the world has to follow, the way Japan exports its bullet-train standards along with the trains.

The Full Story

The bet that paid off: build engineers first, products later

Ather’s origin story is almost stubborn. Mehta and his co-founder Swapnil came out of IIT Madras with one conviction — that they could build things — and not much else. The product instincts came later. The engineering confidence came first. They took the conviction so literally that the first 300 hires were all engineers by training, including the people doing HR, sales, and creative work.

“We can build anything that anywhere else can be built in the world.”

This isn’t just a nice founding myth. It became a filter for strategy. If a business move doesn’t lean on an engineering advantage, Ather is instinctively suspicious of it — not because it’s a bad idea, but because it’s not their kind of idea.

“Why should Ather pursue something if there is no engineering and product moat behind the entire approach?"

"Leverage” — the engine that builds engines

The word on Mehta’s mind lately is leverage. Think of it like this: for years, every new product was a blank sheet of paper. Want to build a new thing? Start from zero, fumble around, learn how. After 13 years they’ve finally built what he calls an organizational engine — a repeatable pipeline that takes a founder’s rough pitch and reliably turns it into concept, then prototype, then a fully industrialized product, year after year.

“As a founder, we can imagine a product and there is an entire engine laid out that can convert that literal pitch into a concept and then a prototype and then very quickly a fully scaled up industrialized product.”

That’s why the company is suddenly doing a lot of things at once — a helmet, more two-wheelers, international expansion, a standards push. For the first time the ideas aren’t overwhelming; there’s a machine waiting to absorb them.

Vertical integration, redefined

Here’s the most interesting reframe in the conversation. In India, Mehta says, “vertical integration” almost always means manufacturing integration — you don’t just assemble the vehicle, you own the paint shop, then the raw materials, eventually the mine. Backward, backward, backward. It’s a hundred years of a country that has only ever been allowed to be a factory, never a design house.

“I feel like in India we use vertical integration for one reason only, which is manufacturing integration. And I think it kind of betrays a fairly old kind of thinking.”

His version: integrate the design, not the dirt. Own the IP. Bring the mechanicals, electricals, software, and design under one roof so the winners are the companies that “go straight to the silicon and design from that point onwards.” Manufacturing isn’t changing fast anymore. Power electronics, embedded hardware, batteries, software — those are. That’s where the value is migrating, so that’s what you own.

The concrete payoff: when one company controls the whole stack, features that used to be locked behind cost and complexity suddenly fall out almost for free. Going electric isn’t just swapping a petrol engine for a motor — it’s that your control over the whole vehicle “skyrockets.”

“We can build traction control at a 100 rupee cost increase because everything is there.”

His sharpest example: anti-lock braking (ABS). There’s been pushback against mandating it on Indian two-wheelers because it adds about ₹10,000 to a petrol scooter. On a properly architected EV, he claims, you can deliver ABS-like behaviour for roughly ₹100 — because you already have a motor controller running fast control loops; you just repurpose it. Features that were “gate-kept” — traction control, cruise control, cooling — get unlocked.

The screen nobody asked for

Twelve years ago Ather made a strange bet: put real compute and a 7-inch touchscreen on a two-wheeler, before anyone knew what it was for. The only use case they could honestly imagine was Google Maps.

“The only use case that we could imagine, honestly, hand on heart, was Google Maps.”

It didn’t work well at first — 2G connectivity, underpowered hardware, everything laggy. But they were selling premium scooters to high-end buyers, so a half-working dashboard was “a sprinkling on top of a cake.” The discipline was to keep shipping it regardless, purely to justify continuing the software work behind it. Competitors then had to match the spec, and the whole Indian two-wheeler market ratcheted upward. A global supplier of two-wheeler displays eventually told them India had run ahead of what the rest of the world was even developing.

The next twist on the same idea: smart helmets and on-device or cloud LLMs that could make the dashboard itself irrelevant — a heads-up display on the visor plus voice, so you never look down. Not legal yet, but plausible now in a way it wasn’t before. He also notes Indian consumers are quietly willing to pay for always-connected features, and the cost of providing them is far below what people will pay.

The same logic explains why Ather built on Android from day zero — an open platform they could run apps on and control, rather than handing a fast-moving piece of technology to a supplier who sits two levels removed from the customer.

Batteries became boring (in a good way)

When Ather started, a battery felt like a semiconductor — a proprietary black box, hard to make, priced like a mystery. Today the cells have basically become a commodity priced near their raw-material cost. That’s quietly enormous: it makes the total-cost-of-ownership case for switching from petrol to electric genuinely attractive. The catch is that you don’t control commodity swings — if lithium 10x’d or cobalt 5x’d, that’s out of your hands.

The IP didn’t disappear, though; it moved up — into how you integrate the pack and the algorithms in the battery management system. He credits early architectural calls (like making everything updatable over-the-air) for letting that value accrue. Change those platform decisions later, he warns, and it’s brutally hard.

On battery swapping — pulling a charged pack from a Blinkit-style courier — he’s skeptical for ordinary riders. Every removable interface invites failure: drop a pack a foot while sliding it in and you can break terminals and start a fire; water can seep in and short it. Sealing and armouring against all that adds weight, kills range, and adds cost. It still makes sense for gig-economy riders where fast charging isn’t fast enough, but the bigger obstacle now is product, not engineering.

The standards sermon

This is the part Mehta clearly cares about most, and he widens it beyond Ather. India, he argues, has barely ever set a global technical standard — NPCI (the body behind UPI) being a rare exception. Ather open-sourced its charging standard and got it adopted by India’s standards body (BIS); the code doesn’t sit with Ather, so it’s neutral. Now the ambition is to get the world to adopt the Indian standard.

His warning is the memorable bit. If you pioneer a technology but don’t wrap it in an easy-to-use standard and push that standard abroad, someone else will write the standard the world ends up needing — and then it flips back on you. You’ll have to build to their standard, and the best suppliers for it won’t be Indian.

“When the Japanese export bullet trains, they’re also exporting all these standards that come with it — the glasses, the bolts, the steel. And the best suppliers for that steel will not be an Indian company. It’ll be a Japanese company.”

“Wherever we have taken a lead, solidify the lead by creating a standard tsunami out of it.”

Loose ends: grid, talent, and getting bored

A few smaller threads. As EVs scale, he sees a more decentralized energy economy where electric vehicles actively help balance and stabilize the grid, with things like time-of-day electricity pricing going mainstream. He’s bullish on a revival of unglamorous engineering — manufacturing, tooling, dies, precision metalwork, castings — arguing software architecture jobs are finite while a real industrial nation needs armies of manufacturing engineers.

On talent: in a startup, you only retain great engineers if the founders are genuinely, visibly excited about engineering. On staying fast at scale: decentralize aggressively to the “leaf layer” while founders keep ridiculous direct oversight — talk to the person actually building the dashboard, not the VP above them. And on his own working style, a nice closing note:

“I really like getting bored… you’re really at your thinking best when you have nothing else to think about.”

Key Takeaways

• Ather hired only engineers (by training) for its first ~300 employees — including HR, sales, and creative roles — to hard-code a “we can build anything” culture.
• The strategic filter: reject any strategy that doesn’t rest on an engineering/product moat, even if it’s a good strategy for someone else.
• “Leverage” = an organizational engine that reliably converts a founder’s pitch into an industrialized product. Building the engine took 13 years; the payoff is doing many products at once.
• Reframe of vertical integration: own the design and IP (silicon, power electronics, software, BMS), not the factory and the mine. Manufacturing is slow-changing; the fast-changing layers are where value lives.
• Electrification’s real unlock isn’t the motor — it’s software control over the whole vehicle. Traction control for ~₹100; ABS-like function for ~₹100 vs ~₹10,000 on a petrol bike, by repurposing the motor controller’s fast control loop.
• Ship ahead of demand to justify continued R&D: Ather forced a compute + touchscreen onto scooters years before there was a use case, dragging the whole Indian market upward.
• Build on open platforms (Android from day zero) and keep fast-moving tech in-house, because suppliers sit two levels from the customer and can’t anticipate needs.
• Battery cells have commoditized to near raw-material cost; the IP migrated up into pack integration and battery-management algorithms. Early OTA-update architecture made capturing that value possible.
• Battery swapping for mainstream riders is held back by the removable interface (fire risk on drops, water ingress) and now mainly by product fit — it survives in gig-economy use cases.
• Standards capture is national strategy: pioneer a tech and write/export its standard, or someone else writes the standard the world adopts and it flips back against you (the Japanese bullet-train analogy).
• Staying fast at scale = decentralize to the “leaf layer” but keep founder-level direct oversight, talking straight to the individual contributor.
• Be “really good with numbers and the story” — combining both makes a powerful analyst and a powerful business storyteller.

Claude’s Take

This is a corporate podcast — NPCI interviewing a founder, two genial “Taruns” lobbing soft questions — so calibrate for that. Nobody asks Mehta about Ather’s losses, its valuation, the brutal price war Ola and TVS and Bajaj are running, or whether the “we’ll only do things with an engineering moat” purity has ever cost them. It’s a highlight reel, not a cross-examination.

That said, the thinking is genuinely sharp and worth keeping. Two ideas earn their keep. First, the redefinition of vertical integration from owning the factory to owning the design is a clean, transferable mental model — it applies well beyond scooters, and the ₹100-ABS example makes the abstract concrete. Second, the standards-capture argument is the kind of thing more Indian operators should internalize; the bullet-train framing is memorable and basically correct about how industrial power compounds.

Where to keep a raised eyebrow: the “first 300 hires were all engineers” line is told as triumph, but he himself admits it wasn’t the most correct principle, and the cost of that monoculture is conveniently unexplored. The ₹100 ABS-like number is doing a lot of rhetorical work — note the careful “-like” hedge; it’s a control-loop trick, not certified motorcycle ABS, and he glides past the difference. And “I like getting bored / books just flood your mind” is a charming founder aphorism that I’d file under personality, not advice.

Net: a 7. No real BS, several ideas that will stick, but it’s an admiring interview with no friction, so you’re getting Mehta’s best self rather than a stress test of his claims.