Stellaris Enterprise Companions is raring to again AI startups however conviction is vital: Companion Alok Goyal

Enterprise capital agency Stellaris Enterprise Companions is doubling down on AI and deeptech startups in India. But it surely’s not speeding headlong into the AI frenzy. 

As a substitute, the VC agency is taking its time to seek out its ‘candy spot’ amid the flurry of AI startups and wannabes mushrooming throughout the nation—a vantage level from which it will possibly consider and again use circumstances it believes in. 

Final yr the corporate stated it could make investments $100 million-$150 million on AI startups in India. In accordance with Alok Goyal, Companion at Stellaris, the agency’s focus can be on software enablers throughout the AI ecosystem. “We’re taking a look at all the pieces from agent builders, to guardrails, to safety… We’re extra bullish on these AI use circumstances,” he notes. 

Having stated that, the agency just isn’t pigeonholing itself with regards to its funding bets. It’s not hung up on a selected thesis or an funding goal. It’s open to concepts and able to drift with a founder-first method. 

“There aren’t any guidelines in our enterprise. We are able to have our ideas, after which one meets a improbable founder who’s doing one thing, and our thesis can exit of the window at that time limit,” says Goyal. 

He believes right now’s ecosystem is not only about enterprise AI, and shopper AI is gathering momentum rapidly. 

In January this yr, Stellaris led a $3-million funding spherical in voice AI startup Arrowhead. The agency has additionally invested in B2B advertising and gross sales analytics platform Components.ai and AI-driven intelligence platform OrbitShift.

Whereas it’s enthusiastic about companies constructing vertical AI functions and good providers, Stellaris is comparatively much less bullish on cloud gamers and firms constructing semiconductors and enormous language fashions (LLMs), primarily on account of their larger capital necessities. “There are very massive, however only a few winners (in these sectors),” says Goyal. 

Based in 2016 by Ritesh Banglani, Alok Goyal, and Rahul Chowdhri. Bengaluru-based early-stage VC agency Stellaris Enterprise Companions has backed D2C manufacturers akin to Mamaearth and Nestasia.

In an interview to YourStory, Goyal breaks down how—is navigating the AI increase, its investments plans in deeptech, and the way the agency is viewing early-stage funding. 

Edited excerpts:

YourStory [YS]: What’s your studying of the early-stage startup panorama in India right now?

Alok Goyal [AG]: The early-stage ecosystem was in a hyper-drive in 2025, and 2026 appears to be like like a continuation of that momentum. In distinction, 2024 was slower. The dry powder was there, however traders have been cautious after the 2021 highs and the reset by 2022–23. Add to that the fast rise of AI, individuals have been nonetheless determining the tempo of change.

Now, traders are extra snug. Cheque writing picked up in 2025, particularly at Collection B the place deal rely has reportedly doubled year-on-year. Development investing is inching again, and with public market liquidity bettering, late-stage confidence ought to strengthen in 2026.

Sector-wise, two clear themes stand out to us: shopper tech and AI. AI is not simply enterprise, shopper AI is scaling quick. Total, momentum is broad-based, however later-stage funding is the place the true shift might occur subsequent yr.

YS: Is there a selected layer of the AI stack you’re centered on?

AG: In case you break the AI stack down, it begins with semiconductors—the NVIDIAs constructing coaching and inference chips. Then come cloud gamers like AWS, Google Cloud, and Azure. On high of that sit LLMs, then software enablers and at last apps and providers. Information and compute energy your entire stack.

We’re typically extra bullish on the highest half of the appliance layer, much less on foundational fashions. That house wants heavy capital, deep PhD-level expertise, and tends to provide few, outsized winners. India has robust groups, however not sufficient density but.

The place we’re excited is vertical AI functions. AI permits you to rethink each cloud workflow and the chance is extremely granular. You’ll be able to go as micro as employees’ compensation underwriting in US insurance coverage or private damage regulation name dealing with, and construct massive companies there.

The opposite huge shift: the road between software program and providers is blurring. The most effective corporations aren’t simply promoting instruments, they’re delivering outcomes. AI will get you 80–90% there; a skinny human-in-the-loop layer closes the hole. It’s not BPO-scale providers, however a wise providers wrapper round deep merchandise.

That mannequin performs to India’s strengths: robust product expertise plus cost-efficient service layers. That’s the place we see sturdy alternatives.

Additionally Learn

Insurtech startup Pibit.AI raises $7M led by Stellaris Enterprise Companions

YS: Given the surge in AI startups, what differentiates those you put money into? Is there a transparent framework or is it extra founder-led?

AG: On the stage we put money into, corporations typically pivot, generally dramatically, from what we initially backed. So our lens is deeply founder-first.

We search for edge. Studying agility. Obsession with the issue. And truthfully, a streak of madness—that unreasonable perception that they’ll bend actuality. With out that, it’s very exhausting to construct one thing nice.

Are we all the time proper? After all not. Even nice founders don’t win each race. However we’ve seen sufficient circumstances the place we backed them for X and so they constructed Y, and we’re glad we backed the founder, not simply the thought.

YS: Many VC companies right now run structured cohorts and mentorship programmes for AI founders. Is that one thing you’re contemplating as nicely?

AG: Totally different funds, totally different phases.

There are funds that are available a lot earlier. We are available a bit later, and our mannequin may be very concentrated. We do only a few offers, however after we make investments, we go all in. We intention for significant minority possession and keep deeply concerned. It’s not a spray-and-pray method. It is simply what works for us.

We’ve seven individuals writing cheques. With the newest fund, we do about 10 investments a yr, roughly one to 2 offers per accomplice yearly. Fewer bets, deeper conviction.

YS: With extra AI offers coming by, has your funding cadence shifted in any respect?

AG:  No huge shift. The one motive our deal rely has gone up is capability. We have been three companions, now we’re 4, plus three principals. So seven individuals who can write checks. That’s taken us from 6–7 offers a yr to about 10 now. However we’ll by no means be high-volume traders. 

In deeptech, we’re generalists, and nothing is off the desk. However deal circulate has elevated considerably. Earlier, considered one of us checked out deeptech; now 4 or 5 do. And we don’t deal with it as one bucket anymore—semiconductors, biotech, robotics, spacetech are all very totally different.

We hadn’t diligenced a semiconductor or biotech firm in eight years. Final yr, we did a number of. So, high quality and quantity have gone up.

That stated, we’re disciplined. Deeptech is thrilling, nevertheless it has an extended gestation and is capital-intensive. We’ve to underwrite threat: How far is income? How a lot capital earlier than it exhibits up? Will returns justify the danger?

We’ve accomplished a number of deeptech investments that aren’t introduced but however we don’t make investments simply because one thing is deeptech. It’s all the time bottoms-up: founder, enterprise, risk-reward.

YS: What sort of startups have you ever invested in?

AG: We’ve accomplished one in house and one in biotech, each of that are thrilling areas. Area, specifically, is heating up. India is without doubt one of the few international locations globally that may commercially launch satellites. Outdoors of the US, China and a European bloc, there aren’t many actual choices. That’s a strategic benefit.

We’re seeing robust expertise spin out of ISRO, and even second-generation founders are rising from earlier house startups. That virtuous cycle has begun.

Additionally, satellite tv for pc volumes have exploded. Earlier, launches have been restricted and largely geostationary. Now, low Earth orbit dominates—80–90% of satellites are there. Greater volumes imply new ancillary companies: in-orbit servicing, restore, floor stations, communication layers.

With SpaceX resetting ambition globally, daring Indian founders are stepping up. It’s a genuinely thrilling house.

YS: Is there significant exercise on the intersection of AI and spacetech that’s caught your consideration?

AG: It is very important be aware that AI isn’t a sector, it’s a functionality. Like cloud. Right now, nearly each firm makes use of cloud. AI is changing into the identical. It will get embedded throughout dozens of workflows inside one product.

Take a satellite tv for pc restore mission. You’re approaching a broken satellite tv for pc in low gravity. The ultimate 10–20 metres require excessive precision—tiny propulsion bursts, visible steering, micro-adjustments. One unsuitable transfer and also you harm a multi-million-dollar asset.

AI powers the navigation. Then it helps management the robotic arm that attaches a brand new part. That’s one other layer of intelligence.

So whereas it appears to be like like ‘one’ house firm, there might be 50 totally different AI functions inside that single system. That’s how pervasive AI actually is.

YS: From the $100 million–150 million allocation to AI, are you working in direction of a selected variety of investments?

AG: We don’t run with targets. It’s very bottoms-up for us.

We’re conscious to not get overly skewed in a single path, however past that, we’re open. It might be shopper AI, enterprise AI, and stage isn’t a constraint both. Most find yourself being seed, generally Collection A.

We don’t sit with a quota in thoughts. If we meet an awesome founder constructing one thing compelling, we again them. Easy as that.

Additionally Learn

Stellaris Enterprise stays centered on concentrated startup investments

YS: Is the cheque measurement in AI broadly in step with your customary funding vary?

AG: Our cheque measurement is pretty broad, and we make investments anyplace between $1 million and $10 million. It is dependent upon stage and valuation. If the value is smart for the possession we wish, we’ll make investments.

Our objective is to be a significant minority shareholder.

Comply with-on investments—together with for AI corporations—come from the identical fund pool. It’s one fund, and each preliminary cheques and follow-ons are allotted from there.

YS: How a lot of the fund has been disbursed up to now?

AG: We don’t disclose that determine. 

YS: Wanting again, have been there any sectors the place your early optimism didn’t translate into outcomes?

AG: Truthfully, it’s hardly ever a clear sector name.

Inside the identical sector, you’ll see some corporations doing exceptionally nicely and others struggling. That’s simply the character of enterprise. It’s much less about this sector labored or that sector didn’t, and extra concerning the particular founder, timing, execution, and product-market match.

You’ll be able to choose nearly any sector and discover each outcomes. That dispersion is regular in our enterprise and it’s not all the time a top-down sector development that determines success.

YS: How are you viewing exit alternatives proper now? Have issues picked up meaningfully?

AG:  Normally, exit markets in India have by no means appeared higher.

They’re nonetheless largely IPO-led quite than M&A-led, however with the IPO window open, even M&A is seeing extra momentum. India hasn’t traditionally had many billion-dollar acquisitions and people are nonetheless uncommon. However within the $100 million–300 million vary, exercise is clearly choosing up.

I feel robust IPO exercise will drive extra acquisitions. As soon as corporations go public, they’ve cheaper entry to capital and excessive development expectations. Some development can be natural, however a significant portion can be inorganic. 

YS: How have your funds carried out? With exits choosing up, are there any metrics you take a look at?

AG: I can share broad contours, for the reason that detailed numbers are carefully held with LPs.

Our 2017 fund is our most dependable efficiency indicator due to its maturity. On this fund, we’re firmly within the high 10% globally, when it comes to efficiency. For context, the top-decile return for 2017 is about 21.8% and we’re nicely north of that. 

Whether or not this places us within the high 5% or high 2%, I can’t say however throughout metrics akin to IRR (inner charge of return), DPI (distributions to paid-in capital) and money returned, it has been a big performer.

The newer funds are too early to benchmark meaningfully. We’re a seed-stage investor, and Fund II solely completed deploying about 15 months in the past. It’ll take a pair extra years earlier than comparisons are honest.

YS: For Fund III, what’s the anticipated deployment timeline?

AG: We genuinely don’t intention for a quantity.

It might sound evasive, nevertheless it’s not. We’re utterly bottoms-up and really feel zero stress to deploy. Personally, I didn’t write a single test in 2024. I evaluated loads however simply didn’t have sufficient conviction. Final yr, I did three.

Nobody tracks that internally. Nobody asks why you didn’t make investments. What issues is conviction.

Some months we could do two offers, some years perhaps none individually. Throughout seven traders, it averages to about 10 a yr.

Sometimes, over a 3–4 yr window, we construct the portfolio after which concentrate on follow-ons.