In search of exceptional people in deployed hardware/robotics. Invest early & help founders hire @aforevc @basanitecap | 🇺🇸 🇧🇩 🇭🇰
San Francisco, CA
Joined May 2015
- Tweets2.3K
- Following4.1K
- Followers3.1K
- Likes18.7K
Pinned Tweet
I moved from New York to San Francisco after making the decision in just two weeks (thanks to @vincentzhu for planting the seed last year, @anamitra for pushing me, and Jums for the support)
It was a huge, scary leap months ago, but it’s been nothing short of phenomenal since
Now I’d love to help someone else make the move
If you’re an engineer/technical leader in NY who wants to start a company in SF, DM me. I’ll be in NY from Sat, Oct 31 through Fri, Nov 6 -- I’d love to meet, hear what you’re thinking about building, and share anything useful from my own move
And if you’re not ready to found a company but are considering joining an early-stage startup in SF, reach out anyway
The energy and ambition in SF is unmatched right now. If you’re curious about being part of it, let’s jam!
Startup investing hierarchy:
- Be in the right company
- Own enough of it to matter for your fund size
- Pay a sensible price
And it's hard to skip a step. Ownership in the wrong company doesn’t work, and a great price on a sub-scale position doesn't move the needle either
The way startups are building nowadays is different from a few years ago. Founders ran the lean playbook and do the customer discovery -> MVP -> iterate on usage process
But that only worked when building was hard and demand was unknown. Imo demand is 'easy' to discover nowadays, but the harder part is to build trust that you can deliver the result or build the thing itself
We're back to the derisk the technical 'white-hot risk' playbook from vintage Tom Perkins / Genentech
Today, I think the first milestone is actually more of an experiment vs. a product. Name the technical breakthrough and proof points to get there (this is higher risk but whatever, we're all risk on! for now...)
The good news is that the loop is 10x faster than in 1970s. Simulation, rented compute, contract manufacturing etc. can help you prove things out faster
So more VCs should look like @deedydas to evaluate these product risks, and not just read usage patterns/old playbooks. If you're more of the latter, you better go earlier or build a raison d'être. In this asset class, the company needs to choose you as much as you pick them!
There was a specific point in my life when I decided to be kind, even at the cost of being outwardly 'nice'
Passing on a call/meeting is hard, esp if you're younger or a people pleaser (VCs self-select for this) -- but it's actually kinder to *not* waste someone's time
investors dont pass on founders in calls enough
when i was working on my previous companies i remembered every pass every hedge every follow up not acted on every ghosting
so i strived to not be that and then realized that i can also be even more transparent
so for a few years now ive started to pass on founders mid call (10-15 mins in) or at the end of the call if it's not a fit
saves both sides a lot of time to move on with the rest of the day
Growing up in Bangladesh, Myanmar, and Hong Kong, I got to see first hand how better financial services can change lives for the better
Despite all the hype around stablecoins, penetration is still low. There are good teams in the space and there are intense teams, but then there’s @moneymamba and team who take it to a different level. Every single interaction feels like an all-out attack (iykyk)
I’m so excited and proud to play a role in this team’s relentless execution to connect the global financial system. These hidden taxes add up - and it’s about time we give people better options
Fin.com. $20M Raised. 7 M&A’s completed in 12 months.
Investors include Expa, Coinbase Ventures, and Tenet Fund. 825 million users impacted. You have probably used us already.
We help your favorite fintechs move money to the rest of the world.
This tweet has come up a few times in convos so figured I'd share my take. This really is an issue between optimizing for access or investment returns:
(1) @paulg sees everything, or close to everything, in and around the YC ecosystem. And he probably has amazing follow-on leverage to put more $ to work + pre-existing exposure via YC. If you're a typical angel writing maybe 10-20 checks over a few years, it's probably not the same game...
(2) Your check is too small to determine the next round, but not too small to determine your investment return. $50K uncapped converting at $10M post is 0.5%, and at $100M post (not uncommon after YC nowadays) is 0.05%. Even with the power law, the math ends up mattering esp. if you can get liquidity via secondaries in the hot Series B/C rounds
Now if you don't have pre-existing access, or you think you can 'monetize the relationship' by selling founders a service or something else, maybe it makes sense as CAC with upside. But from a returns perspective, it still feels tough...
In defense of the uncapped MFN SAFE:
As a small angel, surely the optimal strategy is to just write uncapped MFNs all the time. It’s no secret, or surprise, that this is the exact strategy of @Paulg Graham of @YCombinator, undoubtedly one of the best in the game.
I can’t think of a scenario where that isn’t the optimal game theoretic strategy. On the one hand, your cheque is almost certainly too small to impact the price it converts at. And on the other, if you insist on a cap, you may miss the best companies, which in a power law game, is THE cardinal sin.
If anything, the MFN is better than the cap since at least you’re protected if they end up raising more later on SAFEs with a lower cap.
Anyone have a good reason to insist on a cap as a small angel (or small preseed fund)?
[Not] at Mercy of Multiples
The simplest way to break down a co’s value is revenue x market multiple. Venture DD obsesses over the former since there are clear frameworks for it, but the latter gets treated like weather - something outside your control.
Except every check is already a ‘long on expansion’ bet. Paying seed valuations today means you are never neutral to the multiple…
While model capabilities create bigger/faster displacements than ever, history offers clues. SaaS re-priced off license comps, fintech compressed to lending comps, coding tools expanded overnight on a capability threshold. And re-ratings have observable drivers: a tech breakthrough, monopolistic characteristics emerging, early signs of scale/brand accruing in a bake-off etc.
With so many lazily comped categories, there’s opportunity for unique POVs around distribution (can vertical saas players deploy robotics better than the latter going direct?), TAM (labor arb, fintech attach), etc.
Much of the market is priced on an assumption of stagnation: this is what the category is, and this is what it trades at. But multiples don’t drift, they are often driven, and ‘dead’ categories are full of competent teams actively working the re-rating. While most early investments are long expansion by default, I’d rather be long agency by design.
More people - albeit with Sarah’s credibility - should do this
My dream list to meet if anyone’s feeling generous 😉
- William Hockey (product/infra judgment)
- Rich Paul (parallels between athletes and founders)
- Russ Hagey (helping founders transition to scaled CEOs)
Always wondered what’d happen if you took the domain away…
The QSBS rollover is a great capital raising hack for emerging managers. With access to a sufficient volume of priced rounds and tranched closes, I actually think this can turn into a repeatable capital formation strategy
If you’re moved by Cabo Verde’s world cup run, pay a visit to the country!
There’s a direct flight from Rhode Island to Cabo Verde (~7 hours) - a boost to the tourism industry would be a fitting tribute
Love these guys!
We're excited to finally share Vicus Ventures Fund 1 publicly: our debut $55M seed fund to back outlier founders at the earliest stages.
Founded by brothers Raj and Sunny Singh Sandhu, our ethos has been deeply rooted in family.
Fortune Article: fortune.com/2026/06/30/the-v…
