About — Hari Raghavan

Hi! I'm Hari Raghavan. I’ve worn a bunch of hats in the last couple of decades but I really love doing three things:

I love building elegant software
I love investing in great companies (and seeing number go up)
I love writing long-form pieces

I'm a loyalist to some people, some causes, and all dogs.

Work

I’m currently building Underline — a growth stage venture fund, with a twist.

Previously: I founded Autograph and worked on it from 2023 to 2026 — an HRIS → agents for Finance & HR teams. We never quite found PMF but built some powerful stuff we’re going to open source soon. Before that I founded and ran AbstractOps from 2019 - 2023 — a back-office-in-a-box → pivoted to state compliance which is a profitable and growing business today; I stepped into a Chairman role in 2023 to start Autograph.

In a past life, I was part of the founding team and served as the COO for Forge, the largest secondary market platform. The company later went public and was recently acquired by Charles Schwab.

Investing

Along the way, my wife and I did a bunch of angel investing personally — ranging from Seed and Series A checks into Notion, Rippling, Mercury, Invisible Technologies, and a couple of other unicorns / soon-to-be-unicorns. We also invested into a bunch of late stage / pre-IPO companies (something I knew well due to my time at Forge) — Slack, Rubrik, Airbnb, Robinhood, SpaceX, Stripe, Ramp, Mercury — which is collectively at a 49% IRR as of June, and 85% compounded in the 2023-26 cohort. (You can see why I decided to go into investing full-time; my wife is probably correct that I may be a better investor than a founder.)

Between 2021-25, I ran our early stage investing through the Autopilot Fund. We invested in 100+ companies, including several breakouts where we invested at Seed like Wispr (Flow), GC AI, OneImaging, Mastra, Zoo.dev, Tiny Health, Copilot Money, Consensus, NationGraph, Cytronic, Sunflower Sobriety, Valur, Power, Sunset HQ, Rise Data Labs — and more.

Trivia

Some people love food, or art. I love software. And I especially love building it. I literally get goosebumps from a delightful design or magical interaction.
However, I recently came to the conclusion that I .
I’m not very good at being iterative with products or picking a narrow wedge; I love conceiving of the elegant system that solves a wide swath of problems rather than the specific thing that solves the most burning user need. Despite consciously resisting the urge I ended up building extensive experiments instead of the tried-and-true “sell before you build” approach.
I am unlikely to be great at enterprise sales. The main reason for this is a lack of empathy for the typical buyer. A good friend helped me understand “you’re not selling a solution to a company; you’re selling a promotion to your sponsor.” This sort of trade provoked a visceral negative reaction in me, which I understood perhaps too late. But better late than never.
There are lots of other kinds of sales I’m great at — selling a VC during fundraising, selling a candidate during recruiting, selling a financial product to a buyer, or selling a founder on me as an investor. I’ve accepted that you should dedicate yourself to the category where you’re most likely to be successful as a seller, which means I should do those things (and not enterprise sales).
So I try to build things that I think are well designed or fun or useful:

I learned C++ in high school, and built complex “applications” in Excel during my consulting days, but I didn’t pursue a traditional engineering path (despite loving systems design and database architecture). Coding agents today have been incredibly empowering, and over the last ~year and I’ve loved every minute of it.
I was born in the US but grew up in India. I came back here for college. I was a huge FOB and spent a lot of energy intentionally assimilating.
One of my most impactful decisions was to begin writing regularly at the start of 2025. In the ~20 months since, I’ve published more than 80,000 words, mostly in the form of long-form writing (on X arcticles, sometimes crossposted on Substack)
I have extremely confusing political views because I’m a radical moderate. Fun fact: a sitting member of congress (MTG) proposed a civil war over one of my tweets.
I love my dog more than I love almost all people. Get mad at that if you want.
I love fantasy novels — I’m a huge Brandon Sanderson dork.
I’ve seen more than 200 TV shows.
I love good wine, beer, and cocktails, and I think the declining alcohol consumption amongst Gen Z is a bad thing for society.

⚙️

This toggle block is hidden on your Super site

Pages ↓

Content Database ↓

I’m not very good at being iterative with products or picking a narrow wedge; I love conceiving of the elegant system that solves a wide swath of problems rather than the specific thing that solves the most burning user need. Despite consciously resisting the urge I ended up building extensive experiments instead of the tried-and-true “sell before you build” approach.
I am unlikely to be great at enterprise sales. The main reason for this is a lack of empathy for the typical buyer. A good friend helped me understand “you’re not selling a solution to a company; you’re selling a promotion to your sponsor.” This sort of trade provoked a visceral negative reaction in me, which I understood perhaps too late. But better late than never.
There are lots of other kinds of sales I’m great at — selling a VC during fundraising, selling a candidate during recruiting, selling a financial product to a buyer, or selling a founder on me as an investor. I’ve accepted that you should dedicate yourself to the category where you’re most likely to be successful as a seller, which means I should do those things (and not enterprise sales).