What we invest in
For two decades the cheapest thing to build attracted the most capital. That trade is now crowded and the returns show it. We think the next cycle belongs to companies where the technology itself is the barrier — and to investors willing to underwrite that risk at the point where it is still uncomfortable.
The argument
01
Engineering risk is the last real moat.
Distribution advantages erode, brand advantages get bought, and anything a competent team can build in a quarter is not a barrier. What still holds is the thing that took five years and a hundred failed attempts to get working. We would rather own that risk than avoid it.
02
India is finally a place to build hard things.
Design talent that used to be exported now stays. Policy support for semiconductors and clean energy is real money rather than a press release. The missing piece has been domestic capital patient enough to sit through a tape-out or a pilot plant — and that gap is precisely where a family office can be useful.
03
The early cheque is where the asymmetry lives.
By the time technology risk has been retired, the price reflects it. We are a small family office; we cannot win a competitive Series B, and we have no interest in trying. What we can do is show up at idea or pre-product stage, form a view faster than a committee can, and be genuinely additive from that point on.
04
And being small lets us think differently.
A large institution has to justify every position to someone else, which quietly pushes it toward whatever is easiest to defend in a meeting. We answer only to ourselves. That does not make us more cautious — the opposite. It means we can back a company that would never survive an investment committee, sit with it through a development cycle, and be perfectly comfortable being the only name on the page who saw it early.
The four sectors
4 sectors, one question.
These are the only places Oncor writes cheques. With two or three investments a year, focus is not a preference — it is the only way a book this size is worth anything. A company outside these sectors gets a fast, friendly no: not because it is a bad company, but because we would be bad investors in it.
Chips are the hardest thing to build and the hardest thing to displace. India now has the policy support, the design talent and the customer proximity to produce real fabless companies — and almost none of the domestic capital is willing to wait through a tape-out.
What we look for
- Fabless design teams with tape-out-proven leadership
- Edge inference, analogue and power management
- Companies with a design win before a product line
The cheque
Cheque size
On a case-by-case basis
Stage
Idea stage through post-revenue
Geography
India, based in New Delhi
Decision
Days, not months
A good fit for us
- You are building in semiconductors, deep tech, energy or AI
- There is a real technical moat, not a wrapper
- A technical or deeply domain-expert founder is on the team
- Any stage — idea, pre-product, or post-revenue
Not a fit
- Thin application layers on top of someone else's model
- Businesses whose only advantage is being first to market
- Real estate funds, syndications or SPVs
- Services businesses with a software wrapper
How we operate
We are not afraid of early
Idea stage, pre-product, pre-revenue — none of those are disqualifying. We would rather take real technology risk with a founder we believe in than pay up for a company whose risk has already been priced out.
Stage-agnostic, conviction-led
We write cheques from idea through post-revenue. What stays constant is the bar: a technical moat we can articulate, and a founder we would back again. What varies is the stage at which we find it.
The network is the value add
We are a family office, not a fund with a platform team. What we do have is deep operating relationships in the industries we invest in, and an active line into New Delhi's VC funds, family offices and institutional investors. We make those introductions from day one, not at the next round.
We move at founder speed
No committee, no memo circulating for three weeks, no waiting on a partner meeting. We can form a view in days and tell you exactly where you stand and why — so you always know what you are working with.