Khosla Ventures is betting $10M on Ian Crosby, whose last startup, Bench, imploded

TL;DR

Khosla Ventures has invested $10 million in Ian Crosby’s new startup, Synthetic, which seeks to create an AI-driven bookkeeping service. Crosby’s previous startup, Bench, collapsed in 2024, but the investor sees potential in his new venture despite technical uncertainties.

Khosla Ventures has invested $10 million in Ian Crosby’s new startup, Synthetic, which aims to develop a fully autonomous AI bookkeeper. This marks a significant comeback for Crosby after his previous startup, Bench, imploded in 2024. The investment underscores Khosla’s confidence in Crosby’s vision despite past challenges and current technical uncertainties.

Ian Crosby’s new company, Synthetic, is still in the design phase, with the goal of creating an AI system capable of generating accrual-based financial statements without human input. The startup’s prototype currently works for a narrow user group, and Crosby admits it is uncertain whether the technology can scale broadly. The $10 million seed round was led by Khosla Ventures, with participation from Basis Set Ventures and Shopify CEO Tobias Lütke.

Despite Crosby’s previous startup, Bench, collapsing in 2024, Khosla partner Jon Chu explained that the firm often invests in entrepreneurs with controversial or challenging histories, citing Parker Conrad’s journey from Zenefits to Rippling as an example. Crosby’s departure from Bench in 2021 followed disagreements with the board over strategic direction and financial issues. Crosby maintains he was not responsible for the company’s downfall and has since gained experience through subsequent roles at Shopify and founding Teal, which was acquired by Mercury.

Why It Matters

This investment highlights a willingness among venture capitalists to back entrepreneurs with troubled pasts, especially when they pursue innovative, high-risk AI solutions. The focus on fully autonomous bookkeeping could disrupt traditional accounting services, but technical feasibility remains uncertain. The funding also signals confidence in Crosby’s renewed approach and the broader trend toward AI-driven financial automation.

AI Bookkeeping Revolution: 2026 Automation Blueprint for Small Business: Cut Accounting Costs by 50%, Save 200+ Hours Annually, and Master ChatGPT, ... With Zero Coding Skills (The Reclaim Series)

AI Bookkeeping Revolution: 2026 Automation Blueprint for Small Business: Cut Accounting Costs by 50%, Save 200+ Hours Annually, and Master ChatGPT, … With Zero Coding Skills (The Reclaim Series)

As an affiliate, we earn on qualifying purchases.

As an affiliate, we earn on qualifying purchases.

Background

Ian Crosby’s previous startup, Bench, was a prominent online accounting firm that struggled with cash flow issues and strategic disagreements before being acquired in 2024 in a distressed sale. Crosby’s departure from Bench followed a failed acquisition attempt by Brex and internal management conflicts. His subsequent roles at Shopify and founding Teal, which was later acquired, provided him with experience that Khosla Ventures considers valuable for his new venture. The AI bookkeeping industry is currently dominated by companies relying on human oversight, making Synthetic’s fully autonomous goal a notable departure.

“I tend to run towards controversy a little bit. In controversy, groupthink often shapes the narrative rather than the truth of the story itself.”

— Jon Chu, Khosla Ventures partner

“We’re not going to release anything that’s not fully autonomous. It’s that or bust.”

— Ian Crosby

Amazon

autonomous financial statement generator

As an affiliate, we earn on qualifying purchases.

As an affiliate, we earn on qualifying purchases.

What Remains Unclear

It remains unclear whether Synthetic’s AI system will be able to reliably handle the complexity of broader bookkeeping tasks at scale. Crosby admits the current prototype is limited and that foundational models still make significant errors. The technical feasibility of fully autonomous AI accounting is still unproven, and the company’s ability to commercialize at scale is uncertain.

Smart Accounting with AI for Small Businesses: A Complete Guide to Using AI Tools

Smart Accounting with AI for Small Businesses: A Complete Guide to Using AI Tools

As an affiliate, we earn on qualifying purchases.

As an affiliate, we earn on qualifying purchases.

What’s Next

Next steps include further development of Synthetic’s prototype, testing with additional users, and efforts to improve AI accuracy. The company may seek additional funding to expand its capabilities and test scalability. Monitoring how Crosby’s team addresses technical challenges will be key to assessing the startup’s potential.

QuickBooks Online Mastery - The Accountancy Academy (Est. 2000): The Complete AI-Powered Guide and Toolkit to Bookkeeping, Automation & Financial Reporting for E-Commerce and Small Business Owners

QuickBooks Online Mastery – The Accountancy Academy (Est. 2000): The Complete AI-Powered Guide and Toolkit to Bookkeeping, Automation & Financial Reporting for E-Commerce and Small Business Owners

As an affiliate, we earn on qualifying purchases.

As an affiliate, we earn on qualifying purchases.

Key Questions

What is Synthetic’s main goal?

Synthetic aims to develop a fully autonomous AI system capable of performing bookkeeping and generating financial statements without human intervention.

Why did Khosla Ventures invest in Crosby’s new startup?

Khosla partner Jon Chu explained that the firm often invests in entrepreneurs with controversial histories, believing they can learn from past mistakes and succeed with innovative ideas like Synthetic’s AI bookkeeping.

What challenges does Synthetic face?

The main challenge is whether the current AI models can reliably perform complex bookkeeping tasks at scale, as Crosby’s prototype is still limited and prone to errors.

How does Crosby’s past influence this investment?

While Crosby’s previous startup, Bench, collapsed, his subsequent roles and experience have convinced Khosla Ventures that he has learned from past mistakes and is capable of leading an innovative AI-driven project.

You May Also Like

Building ML framework with Rust and Category Theory

A working draft explores building a machine learning system using Rust and category theory, emphasizing structured, maintainable pipelines.

It takes two neurons to ride a bicycle (2004)

Researchers demonstrate a two-neuron network capable of controlling a virtual bicycle, challenging previous assumptions about complexity needed for such tasks.

Predictive Maintenance for Mining Rigs Using Machine Learning

Boost your mining rig efficiency with machine learning-driven predictive maintenance—discover how to prevent failures before they happen.

Autonomous Economic Agents: Bots Negotiating on Your Behalf

Smart autonomous economic agents negotiate on your behalf, transforming markets—and understanding their mechanics could be the key to leveraging their full potential.