Ori Eldarov
@leveredvlad
CEO @ OffDeal - building the world's first AI-native investment bank
Last year I raised $5M to build a brand new investment bank from scratch. Not software for banks or a fintech marketplace - a real M&A advisory firm. This is a very long post explaining what we're doing, why, and how it's going (sorry, I don't like threads). OffDeal: The First…
You don't need to spend 6 years tweaking logos in powerpoints to become a great dealmaker.
What finance related opinion do you have that will have people like this?
also true for roofers, plumbers, pest exterminators etc. america has literally millions of amazing operators, most of whom would swim circles around the median tech / finance bro in the real world.
tech nerds in sf have such a narrow idea of entrepreneurship there are so many disciplines that mimic the same business skillset as a startup music artists, content creators, dropshippers, shitposters, standup comedians, onlyfans models, and authors are all founders too
One under discussed point is that since this is a web-app, people on a Mac can finally have a proper Excel experience!
Shortcut – the first superhuman excel agent – is live. While not perfect, Shortcut beats first year analysts from McKinsey/Goldman head-to-head 89.1% (220:27) when blindly judged by their managers. We even gave humans 10x more time. Try Shortcut now (before your boss does).
Many such cases in investment banking as well. Had several search funds submit LOIs / provide markups that were clearly AI generated... I get the desire to save some money, but if left unchecked it might cost way more down the line..
Sam Altman just confirmed what lawyers have been saying for over a year: There’s no legal privilege when you use ChatGPT. So if you’re pasting in contracts, asking legal questions, or asking it for strategy, you’re not getting legal advice. You’re generating discoverable…
This oddly also applies to people that want to be founders.
A lot of young people want to be great investors but refuse to narrow their scope They are constantly chasing the newest trend: semis, AI, crypto, quantum Just remember that trying to become an expert in everything is the same as being good at nothing Here is my unsolicited…
I tell juniors all the time - urgency and responsiveness are the two most important things in finance. Treat things that need a detailed and fast response seriously and you'll stand out before it becomes table stakes.
Paradoxically, the flurry of buyer outreach makes owners more likely to engage a banker - some because they're overwhelmed, others because they want buyers to compete with one another. Makes our job so much easier when there's a pull!
All the proprietary outreach happening from Searchers and PE is only making business owners even MORE inflated in their egos and asking price expectations since they all think they're special now since buyers are burning down their doors every day begging them to sell to them.
The main reason I write is to generate new ideas and connect the dots. Every time I’ve written a blog post, the final product looked very different than what I initially had in mind. AI violates this - it encourages the writer to skip the ‘rumination’ step, in most cases.
Yes. Writing is not a second thing that happens after thinking. The act of writing is an act of thinking. Writing *is* thinking. Students, academics, and anyone else who outsources their writing to LLMs will find their screens full of words and their minds emptied of thought.
100% true. A great CIM is not just a pretty pack of slides. It has to provide an accurate and nuanced view of the business and sell the "story" to investors. A pretty CIM with bad content will just land you 100s of DDQs from buyers (at best).
The attention to detail on the content. With most CIMs, I have about 30-40 follow up questions. They do an insane job of addressing all of them up front, which literally just makes everything faster. Less emails back and forth. It is just easier.
Aaron is exactly right here. Most people (especially outside tech) have the expectation that AI has to be perfect to be really useful - and when it isn’t they give up on it and say ‘the tech is not ready’. Those will people will be left behind (if they don’t adapt).
Aaron Levie says the future of work has flipped. We thought AI would fix our mistakes. Now it makes the first draft. We QA what comes back. “The human's job is to fix the AI errors, and that's the new way we are going to work.”
Windsurf saga has been so insane over last 72 hours. I wonder if team used bankers on this, or a bunch of engineers raw dogged such a complicated sell side.
Same is true for investment banking. PE firms and F500 have the skill to sell the business themselves (they're all ex bankers), but they want a scapegoat in case something goes wrong.
People fundamentally misunderstand the value of hiring consultants Even before AI you could have hired a college grad to do research and find industry data points The real value of hiring consultants is being able to point fingers at someone in case something goes wrong
Interviewed a candidate from GS IB team who flagged that he didn't like the "culture" there. He said he routinely works till 3-4am most weeknights, and then usually both days on the weekend. The worst part was not the hours or volume of work - but waiting until 11pm to get…
many such cases
Feels like an entire generation is studying and gunning for jobs that will mostly not exist or be materially different in a few decades
is it just me or o3 gives WAY too many tables. no matter what question you ask it, every answer is presented as a giant table - is this what user feedback showed OpenAI as the most optimal UX? it pushed me to use Claude/Gemini more because of it
o3-pro is awesome for super complex problems, but i wish they exposed more of the "reasoning" thoughts the way regular o3 / Gemini 2.5 pro does. first, it's just better UX than waiting 5-10 minutes, but also helpful to understand how the model arrived at its answer.
Parker Conrad is exactly right - it is much easier to replace ops with software when you're building the org de novo. That's why we went with our "full stack startup" approach and why we've been able to automate/ optimize so much with AI.
.@parkerconrad on "AI rollups" and how he developed an allergy to ops at Zenefits: "it's very hard to replace ops with software after you've scaled it up"
We recently made a (risky) decision to switch to memo-style CIMs to optimize them for LLMs. The feedback from our PE buyers has been phenomenal - they love the new format and the # of DDQs dropped by ~80%. And we can now make a CIM in <60 minutes. This is the way.
I’ve been converted. I never want to make or read another PPT deck again. Written memos with applicable charts is the only way to communicate ideas.