(AI) INVESTMENT BANKING

The “Artificial Investment Banker” for Small-Business Deals

OffDeal, a startup that uses AI to speed up small-business M&A, has raised $12 million in Series A funding led by Radical Ventures, bringing its total capital to $17 million.

OffDeal is betting that AI can automate the front-end of M&A from buyer discovery to first contact. The New York-based startup wants to cut the time and headcount it takes to run a sell-side process. The company says that with a pitch deck or a website link, its AI can

• Scrape millions of private-company records to build a target list

• Run light diligence to narrow that list down

• Draft and send personalized emails to potential buyers at scale

Follow-ups are automated. Human dealmakers still run strategy, valuation, and negotiation.

The company started by working with search funds, or individuals who raise money from investors to buy and run a small business, and says about a dozen were using the platform during its private testing phase. By the middle of this year, OffDeal had run more than 30 sell-side mandates.

The economics are closer to tech than traditional advisory: a success fee of 5%–10%, with AI doing much of the labor.

Radical Ventures led both rounds:

• $12M Series A in July 2025, joined by YC, Rebel Fund, and Centre Street Partners

• $4.7M seed in September 2024

Takeaway: OffDeal shows how AI-driven buyer discovery and automated outreach can shrink the cost of small-business M&A, opening an underserved part of the market since humans alone couldn’t make money on these transactions.

Related:

No AI? No exit, says investment bank Artis Partners (Finextra)
Fintechs that fail to embed AI into their operating model will find it increasingly difficult to orchestrate an exit.

FUNDRAISING

Excel-Based AI Modeling Gains Momentum

LLMs, by themselves, are bad at math. If you asked ChatGPT two years ago to do simple arithmetic, like 241 - (-241) +1, you’d get a wrong answer.

LLMs are optimized to predict the next word, which makes them better suited to conversational tasks than to structured spreadsheet work. As a result, most AI-for-finance startups concentrate on chatbot-driven equity research, and relatively few aim to simplify Excel.

But now, LLMs generally know when to hand off the arithmetic to a code interpreter. Try it yourself.

That shift has opened the door to new products that blend natural language input with spreadsheet precision.

Recently two companies, Fundamental Research Labs and Endex.ai, have raised cash for agentic Excel.

• Fundamental Research Labs closed a $33 million USD Series A; its product is Shortcut, an Excel-style spreadsheet that lets users ask in plain language to build a leveraged-buyout (LBO) or discounted-cash-flow (DCF) model. (WSJ)

• Endex.ai raised $14 million early stage round led by the OpenAI Startup Fund, and an OpenAI case study shows it using o-series reasoning models to cut latency and improve multi-step accuracy.

Takeaway:

Both startups combine a probabilistic LLM with a deterministic calculator and give users an audit trail. Maybe AI will be the thing that actually shortens investment bankers’ hours, a long-standing issue on Wall Street.

ADOPTION

AI Is Ready; Are the Professionals?

We’re in a strange phase of AI. The technology is clearly good and getting better, but some experts are reluctant to admit it’s effective in their own fields. Some folks are resistant as a point of professional pride.

It’s uncomfortable to acknowledge that a “machine” can do a professional’s job well. My initial reaction to ChatGPT was that it can’t write.

It reminded me of a stat from The Checklist Manifesto, a book by Atul Gawande about reducing failure in complex professions. When surgeons were asked if pre-op checklists were worth the effort, one in five said no. But when asked if they’d want their own surgeon to use a checklist, 93% said yes.

This week, I saw a few stories that echoed that same tension:

• The surprisingly sound financial advice I got from a chatbot (Vox)

• I’m a Therapist. ChatGPT Is Eerily Effective. (Washington Post)

• The rise of AI tools that write about you when you die (Washington Post)

REGULATION

I think of legal industry as somewhat of a proxy for financial services in its adoption of AI. Laws change, regulations shift, but not as quickly as markets.

And you can see from this CNBC headline, AI for lawyers is catching on fast:

Legal AI startup Harvey hits $100 million in annual recurring revenue

• Harvey AI reached $100 million annual recurring revenue (ARR) just 3 years after launching in 2022

• The company provides AI-powered legal platform for lawyers at law firms and corporations, helping with research, drafting, and diligence projects

• Growth driven by expanding usage - over 500 customers including Comcast, with weekly users quadrupling in the past year

Takeaway:

AI faces tougher technical hurdles in finance than in law. Large language models often stumble on numerical precision, and market data can go stale in a single trading day. Yet every expert I’ve interviewed believes these obstacles can be overcome.

WHAT ELSE I’M READING

• Bond Trading Bots Are Getting Smarter and Powering Through Market Unrest (Bloomberg)

• Wells Fargo Deploys AI Agents Business-Wide (PYMNTS)

• Is AI Killing Entry-Level Jobs? Here’s What We Know (Bloomberg)

• UBS took a sweeping look at the AI revolution and concluded the ‘visible’ impact is at least 3 years away for consumer firms (Fortune/Yahoo)

• US SEC Announces AI Task Force to Boost Efficiency (PYMNTS)

• Evident AI Outcomes Report - how 50 banks are using AI (Evident)

• Menos AI Raises $5.2M for AI Idea-Generation Agent (Business Wire)

IN CASE YOU MISSED IT

AI Stack with Harry Mamaysky of QuantStreet

I started a new interview series where I chat with investors on how they're actually using AI that I'm calling AI Stack. My first conversation was with Harry Mamaysky, QuantStreet Capital co-founder and Columbia Business School professor.

From the interview:

𝘐’𝘮 𝘳𝘦𝘭𝘺𝘪𝘯𝘨 𝘰𝘯 𝘪𝘵 𝘮𝘰𝘳𝘦 𝘦𝘷𝘦𝘳𝘺 𝘥𝘢𝘺 𝘵𝘩𝘢𝘯 𝘐 𝘸𝘢𝘴 𝘵𝘩𝘳𝘦𝘦 𝘮𝘰𝘯𝘵𝘩𝘴 𝘢𝘨𝘰. 𝘊𝘰𝘯𝘴𝘵𝘢𝘯𝘵𝘭𝘺, 𝘐’𝘮 𝘢𝘴𝘬𝘪𝘯𝘨, “𝘏𝘰𝘸 𝘤𝘢𝘯 𝘐 𝘶𝘴𝘦 𝘵𝘩𝘪𝘴 𝘵𝘰 𝘣𝘦 𝘮𝘰𝘳𝘦 𝘦𝘧𝘧𝘪𝘤𝘪𝘦𝘯𝘵?” 𝘍𝘰𝘳 𝘦𝘹𝘢𝘮𝘱𝘭𝘦, 𝘵𝘩𝘦 𝘴𝘤𝘳𝘪𝘱𝘵𝘪𝘯𝘨 𝘭𝘢𝘯𝘨𝘶𝘢𝘨𝘦 𝘶𝘯𝘥𝘦𝘳 𝘎𝘰𝘰𝘨𝘭𝘦 𝘚𝘩𝘦𝘦𝘵𝘴 𝘪𝘴 𝘑𝘢𝘷𝘢𝘚𝘤𝘳𝘪𝘱𝘵; 𝘧𝘰𝘳 𝘌𝘹𝘤𝘦𝘭 𝘪𝘵’𝘴 𝘝𝘪𝘴𝘶𝘢𝘭 𝘉𝘢𝘴𝘪𝘤. 𝘐’𝘮 𝘯𝘰𝘵 𝘢𝘯 𝘦𝘹𝘱𝘦𝘳𝘵 𝘪𝘯 𝘦𝘪𝘵𝘩𝘦𝘳. 𝘐𝘯 𝘵𝘩𝘦 𝘱𝘢𝘴𝘵, 𝘪𝘧 𝘐 𝘸𝘢𝘯𝘵𝘦𝘥 𝘌𝘹𝘤𝘦𝘭 𝘵𝘰 𝘥𝘰 𝘰𝘱𝘵𝘪𝘰𝘯 𝘱𝘳𝘪𝘤𝘪𝘯𝘨, 𝘐’𝘥 𝘯𝘦𝘦𝘥 𝘵𝘰 𝘸𝘳𝘪𝘵𝘦 𝘢 𝘉𝘭𝘢𝘤𝘬-𝘚𝘤𝘩𝘰𝘭𝘦𝘴 𝘱𝘳𝘪𝘤𝘦𝘳. 𝘐 𝘬𝘯𝘰𝘸 𝘵𝘩𝘦 𝘮𝘢𝘵𝘩, 𝘣𝘶𝘵 𝘐 𝘥𝘪𝘥𝘯’𝘵 𝘬𝘯𝘰𝘸 𝘵𝘩𝘦 𝘴𝘺𝘯𝘵𝘢𝘹—𝘪𝘵 𝘸𝘢𝘴 𝘵𝘦𝘥𝘪𝘰𝘶𝘴 𝘢𝘯𝘥 𝘤𝘰𝘶𝘭𝘥 𝘵𝘢𝘬𝘦 𝘮𝘦 𝘢𝘯 𝘩𝘰𝘶𝘳 𝘢𝘯𝘥 𝘢 𝘩𝘢𝘭𝘧.

𝘕𝘰𝘸 𝘐 𝘫𝘶𝘴𝘵 𝘨𝘰 𝘵𝘰 𝘎𝘦𝘮𝘪𝘯𝘪: “𝘞𝘳𝘪𝘵𝘦 𝘮𝘦 𝘢 𝘑𝘢𝘷𝘢𝘚𝘤𝘳𝘪𝘱𝘵 𝘮𝘰𝘥𝘶𝘭𝘦 𝘧𝘰𝘳 𝘎𝘰𝘰𝘨𝘭𝘦 𝘚𝘩𝘦𝘦𝘵𝘴 𝘵𝘩𝘢𝘵 𝘥𝘰𝘦𝘴 𝘉𝘭𝘢𝘤𝘬-𝘚𝘤𝘩𝘰𝘭𝘦𝘴 𝘱𝘳𝘪𝘤𝘪𝘯𝘨.” 𝘐𝘵 𝘴𝘱𝘪𝘵𝘴 𝘪𝘵 𝘰𝘶𝘵 𝘪𝘯 𝘴𝘦𝘤𝘰𝘯𝘥𝘴. 𝘐 𝘤𝘰𝘱𝘺, 𝘱𝘢𝘴𝘵𝘦, 𝘢𝘯𝘥 𝘪𝘵 𝘸𝘰𝘳𝘬𝘴. 𝘞𝘩𝘢𝘵 𝘶𝘴𝘦𝘥 𝘵𝘰 𝘵𝘢𝘬𝘦 𝘩𝘰𝘶𝘳𝘴 𝘪𝘴 𝘯𝘰𝘸 𝘪𝘯𝘴𝘵𝘢𝘯𝘵.

Harry walks through how his firm uses machine learning to guide asset allocation, forecast tail risk, and build data-driven portfolios.

Click here for the full Q&A 

This interview is from my Sunday newsletter, where I highlight practical uses cases of AI in investing. If you’d like to start receiving these biweekly editions, change your settings here.

CALENDAR

Upcoming AI + Finance Conferences

I’ve put together a calendar of upcoming AI and finance conferences. Let me know if I’ve missed any and I’ll add them. (just reply to this email.) Thanks!

Bolded bullets show recently added conferences.

• Ai4 2025 – August 11–13, 2025 • Las Vegas

  Multi-industry AI event with a strong financial services track.

• AI in Financial Services (Arena) – Sept 9–10, 2025 • London

  Focused on AI strategy, implementation, and ROI in finance.

• Cornell Financial Engineering Manhattan 2025 Future of Finance & AI Conference – Sept 19, 2025 • New York

  A one-day forum on AI, quantitative finance, and hedge-fund strategies, attracting leading quants and industry practitioners.

• Bloomberg-Columbia ML in Finance Conf – Sept 25, 2025 • New York

  Academic–industry event hosted by Columbia University and Bloomberg, focused on ML applications in finance including asset pricing, market forecasting, and LLM risk.

• GAIIM Conference 2025 – Sept 30, 2025 • New York, NY

  Forum on practical applications of AI in investing, featuring tools for research, valuation, and portfolio workflows.

• AIFin Workshop at ECAI 2025 – October 26, 2025 • Bologna, Italy

  One-day academic workshop on AI/ML in finance, covering trading, risk, fraud, NLP, and regulation.

• AI in Finance 2025 – October 27–30, 2025 • Montréal

  Academic event covering ML in empirical asset pricing and risk.

• ACM ICAIF 2025 – November 15–18, 2025 • Singapore

  Top-tier academic/industry conference on AI in finance and trading.

• AI for Finance – November 24–26, 2025 • Paris

  Artefact’s AI for Finance summit, focused on generative AI, future of finance, digital sovereignty, and regulation 

• NeurIPS Workshop: Generative AI in Finance – Dec. 6/7, 2025 • San Diego One-day academic workshop at NeurIPS focused on generative AI applications in finance, organized by ML researchers.