The numbers Brief Episode map Takeaways Charts Quotes Clips Glossary Go next Transcript

Episode 45 · Deep tech · 50 min

The car as software, the corporate as conglomerate

Bosch's Bengaluru software chief on the car that upgrades like a phone on wheels, the centuries-old corporate that can only transform through insulated microcultures, and why India's deep-tech wave must serve its own market first.

Leadership & org Deep tech & hardware Mobility & EV

Dattatri Salagame

MD & CEO, Bosch Global Software Technologies · with Vishal Krishna

49:37

Said in this episode

▶ 0:24

100/yr

Patents filed by Bosch's India centre

Vishal opens by noting the company files on average 100 patents every year.

▶ 3:01

25 yrs

Age of the Bengaluru software centre

Started by Bosch Germany as its largest tech centre outside Germany, the unit completes 25 years in the year of recording — 'a native software company even 25 years back'.

▶ 10:30

5y → 6mo

Engineering-cycle collapse

A car used to be engineered for five years and owned for twenty; software now gets engineered in six months and matures in agile fashion as it is consumed.

▶ 15:40

300–400k

Workforce too big to transform wholesale

Salagame's case for microcultures: thinking you will transform 300,000–400,000 people around the world is simply not feasible.

▶ 30:20

7–8 of 10

Startups that wow him

Of every ten startups he meets, seven to eight wow him — on relevance, energy and passion, not on the depth of their algorithms.

▶ 37:15

500M

Digital natives forcing adoption

If 500 million young digital natives are using a technology, a company has no choice but to jump in, regardless of its relevance to today's products.

The brief

The argument in sixty seconds

Fresh off a rename — Robert Bosch Engineering and Business Solutions became Bosch Global Software Technologies — Dattatri Salagame argues the new name simply declares what the Bengaluru centre has been for 25 years: a native software company, now spawned into Vietnam, Mexico and Poland. The conversation builds one argument with three faces. The car stops being an asset engineered for five years and owned for twenty; it becomes a mobile phone on wheels, upgraded over the air, carrying classes, wellness checks and commerce across domain lines. The corporation that cannot retrain 300,000–400,000 people transforms instead through insulated microcultures running at fusion velocity — with, he admits, mixed success. And India's deep tech will not scale like GMV marketplaces: it must serve the home market first, with talent as the constraint every CEO will lose sleep over for the next five to eight years.

Worth your time if you are

GCC and corporate-transformation leaders

Deep-tech founders choosing their first market

Mobility and auto product builders

Engineering managers staring at hyper-automation

Episode map

Where the conversation travels

Every block is a chapter, coloured by what it's about. Click any of it to jump straight to that minute on YouTube.

Leadership & orgDeep tech & hardwareMobility & EVConsumer IndiaData & digitisationGCCs & services

01Cold open: tracking Bosch for 15 years 0:00 Vishal introduces a company he has tracked for 15 years — Bosch's software arm, filing about 100 patents a year — and asks what has changed since 2008. 02The rename: a native software company 1:00 Salagame frames the RBEI-to-BGSW rename as identity for a compute-driven decade — a 'native software company even 25 years back', Bosch's largest tech centre outside Germany, now spawned into Vietnam, Mexico and Poland. 03The car as a mobile phone on wheels 3:52 From sensors-on-cars in the early 2000s to a car defined largely by software: digital natives prefer experiences to owning assets, features upgrade over the air, and classes, wellness checks and shopping flow into the cabin. 04Software eating software 8:45 There is no single ubiquitous platform — platforms talk to each other — and the industry that ate media and banking now faces hyper-automation eating software development itself, collapsing five-year cycles to six months. 05Reverse osmosis and the inverted pyramid 12:13 A fresh BTech hire knows the new paradigm better than the 20-year veteran, so onboarding reverses into 'reverse osmosis' — orchestrating a marriage between native modern engineering and rich domain experience. 06Microcultures in a 400-year corporate 14:33 You cannot transform 300,000–400,000 people at once, so corporates split into insulated microcultures with their own mandate and speed; copy-pasting startup culture fails, and the honest target is fusion velocity with mixed success. 07India's deep-tech wave: market first 19:08 Indian deep tech must evolve along the market axis — solving problems grounded in the neighbourhood, not exporting like IT services — with talent as the binding constraint and a scaling curve no GMV-minded investor will recognise. 08Every company is a tech company 24:06 Salagame refuses the B2B/B2C split: marketplace routing algorithms are deep tech, AI screening attacks India's doctor-patient skew, and reach — government, NGO, corporate — is a separate question from what a product is. 09Founders, and the driver with an MVP 28:50 Seven or eight of every ten founders wow him on passion and global benchmarks — and his own driver, armed with the world's cheapest data, runs an asset-light onion-trading MVP straight out of an unread Lean Startup. 10Tulip mania and the tech radar 34:53 On metaverse, blockchain and tulip mania: future technologies are exaggerated by necessity, but hype builds the talent base, and BGSW's tech radar turns capability proof points into scalable applications. 11Fuzziness, compartments, mental health 38:39 Fitness as fundamental as brushing teeth, friendliness with fuzziness, life in detached compartments, de-stigmatised mental-health conversations — and children who 'pass through you' rather than sit as trophies. 12One delivery at a time 44:47 Happiness is playing one delivery at a time, and careers turn on timely nudges — like Dr. B. Gurumurthy, the Carnegie Mellon doctorate who returned to IISc at 33 and pushed Salagame into deep engineering.

Takeaways

Ideas to carry out of this hour

The car becomes a mobile phone on wheels

A car used to be engineered for five years and owned for twenty as a static family asset. Salagame's mobility consumer is a digital native who would rather have experiences than own assets, so the car becomes a software-defined gadget for getting from A to B — bought with an initial feature set, then dynamically upgraded over the air so it gets richer and more personalised through its lifetime. The maker's economics shift from a one-time hardware sale to a lifetime software relationship.

▶ 5:20 ▶ 5:56 ▶ 6:30

Products must cross domain lines, because users already do

A passenger attending a Byju's class on a car screen is education times mobility; a wellness check in the cabin is healthcare; shopping from the seat is vehicle commerce. The digital native walks from iPad to car to office without perceiving a break, so products built in containerised domains break the seamlessness the user actually pays for. Nor is there one ubiquitous platform underneath — it is a string of platforms talking to each other to create the experience.

▶ 7:18 ▶ 8:09 ▶ 9:39

Software is eating software — and the pyramid inverted

In the '80s and '90s software ate media, banking and insurance; now hyper-automation is eating software development itself, collapsing five-year engineering cycles into six months and making two-year fixed toolchains untenable. That inverts the knowledge pyramid: a fresh BTech computer-science hire is native to open source and GitHub, while the 20-year veteran is acquiring the habit. Salagame's remedy is 'reverse osmosis' — orchestrating the marriage between young engineers' native paradigm and veterans' rich domain experience.

▶ 10:30 ▶ 12:32 ▶ 14:15

Transform via insulated microcultures, not 400,000 people

Culturally transforming 300,000–400,000 people worldwide is, in his words, not feasible. Large corporates instead move to a conglomerate format: smaller microcultures insulated from one another, each given mandate, differentiated speed and cultural freedom, built to figure out their next shape rather than to last a hundred years. Copy-pasting a startup format into DNA built over centuries fails — the realistic target is a fusion culture with fusion velocity and mandate, and he is honest that the success rate so far is mixed.

▶ 15:40 ▶ 16:13 ▶ 18:32

Indian deep tech must serve the Indian market first

The strongest axis for deep tech to evolve on is the market: a deep-tech startup built to service some other part of the world merely copies the IT-services model. India has hundreds of domestic problem statements — hyper-local vehicle electrification, AI screening for a skewed doctor-patient ratio, agritech disease detection — and the ecosystem needs a triangulation of large companies, young talent and academia. Because deep tech will not scale the way GMV-chasing marketplaces do, it will develop on a different track from India's popular startup ecosystem and won't suit investors hunting quick scale-ups.

▶ 20:52 ▶ 21:10 ▶ 23:29

Talent is what every CEO loses sleep over for 5–8 years

India offers some of the best talent in the world at scale, and Salagame names talent the one thing every CEO will lose sleep over for the next five to eight years; deep tech specifically needs talent that is promoted, nurtured, retained and given elbow room for hard problems. Meanwhile the world's cheapest data is demolishing formal education as a gatekeeper: his own driver built an asset-light onion-trading business — wholesale-to-retail in tier-2/3 towns, an MVP, plan B and plan C — the lean-startup playbook without ever reading Eric Ries.

▶ 22:03 ▶ 32:17 ▶ 33:25

Track hype on a radar; don't buy the amplitude

Most technologies of the future are highly exaggerated today, and the exaggeration is necessary — it recruits believers and builds a talent base, which he calls minimum hygiene for a country like India. BGSW runs a tech radar, builds capability proof points on a few candidates, then joins the dots into scalable applications. Adoption logic has also flipped: if 500 million young digital natives are using something, you have no choice but to jump in, whatever it does for today's product.

▶ 37:15 ▶ 37:31 ▶ 37:49

The numbers, drawn

What the episode measures

Every figure below was said on air — timestamps included, caveats kept.

Conversation share

portion of the hour spent on each theme

Leadership & org26%

Deep tech & hardware20%

Mobility & EV16%

Consumer India11%

Data & digitisation11%

GCCs & services9%

Computed from the chapter map of this episode.

From five years to six months

months

Lifecycle timescales as stated in conversation: the old automotive cadence versus software that matures in agile cycles after a six-month build.▶ 10:30

Worth keeping

Lines that stay

Imagine car like a mobile phone on wheels. You get your updates every now and then, and your car feature-wise gets richer, more personalised to you as you use it.

— Dattatri Salagame ▶ 5:56

People used to hate the software industry because software ate several other industries. Now here we are, where software is eating software.

— Dattatri Salagame ▶ 10:11

If you ask me one thing on which every CEO is going to lose sleep for the next 5 years, 8 years — it's going to be talent.

— Dattatri Salagame ▶ 22:03

Children pass through you. They are not trophies which you flaunt in your cabinet.

— Dattatri Salagame ▶ 43:56

Clips that travel

Short on time? Start here

Mobility and auto product builders

The mobile phone on wheels

The full software-defined-vehicle thesis: digital natives, over-the-air features, and the car that gets richer with use.

4:43 → 8:45 · 4 min ▶ Watch clip

Engineering leaders and CTOs

Reverse osmosis: the inverted knowledge pyramid

Why the fresh BTech hire onboards the 20-year veteran, and how to marry domain depth with the new paradigm.

12:13 → 14:33 · 2 min ▶ Watch clip

Corporate transformation leaders

Microcultures, not copy-paste startups

The honest playbook for changing a centuries-old company: insulated microcultures, fusion velocity, mixed success.

14:33 → 19:08 · 5 min ▶ Watch clip

Deep-tech founders and investors

India's deep tech will not look like marketplaces

Market first, talent as the binding constraint, and why GMV investors will misread the category.

19:08 → 24:06 · 5 min ▶ Watch clip

Students of aspirational India

The driver who built a startup without B-school

Cheapest data in the world, a wholesale-onion MVP, plan B and plan C — the lean startup without the book.

31:44 → 34:53 · 3 min ▶ Watch clip

Glossary

The jargon, unpacked

GCC

Global Capability Centre — the India-based tech centre of a multinational. BGSW began 25 years ago as Bosch's largest tech centre outside Germany, a paradigm Salagame says few GCCs share.

Software-defined vehicle

A car whose features, differentiation and upgrades live in software delivered over the air across the vehicle's life — Salagame's 'mobile phone on wheels'.

Reverse osmosis

His term for knowledge diffusing from fresh hires — native to open source and GitHub — toward long-tenured engineers, inverting the usual mentoring direction.

Fusion velocity

The hybrid operating speed of a corporate that marries the best practices of startups and corporates instead of copy-pasting a startup format it cannot simulate.

GMV

Gross merchandise value — the marketplace scale metric. Salagame argues deep-tech startups will not scale on GMV terms, so marketplace investors misread the category.

Tech radar

BGSW's structured tracking of emerging technologies: build capability proof points on a few candidates, then join the dots into scalable applications.

Connections

If this resonated, go here next

EP 25 · Deep tech

Full transcript

The whole conversation, searchable

Auto-generated captions, lightly cleaned. Click a timestamp to open that moment on YouTube.