SPRX ads

Big accounting firm: "We're going to need their 200 reports and we're going to need to speak to these 50 people." SPRX: "Connect your systems here. Now go away. We'll be done and ready for your review in 7 days." It's funny. Customers don't really care about the AI or our tech stack. They just care that their taxes are done right, with as little of their time as possible.

The boring stuff wins. We just spent months rebuilding our entire data pipeline. Not the kind of project that gets headlines, just weeks of grinding. We dragged our data pipeline from 90% accuracy to 98%. Cool, that’s just 8%. Yeah, except every single percent after 90 takes 10x more work and is the difference between “eh, close enough and here’s millions in credits”, thanks. Everyone loves shiny AI demos (including myself). But demos don’t build trust. Accuracy and reliability do. Customers stay because the system works every time.

An engineering firm uncovered $754K in R&D credits with under 2 hours of work. I talk to a lot of business owners and finance leaders who assume the credit is not worth the hassle. Not because their teams are not doing qualified work, but because the process has always felt disruptive. ᐧ Long painful interviews. ᐧ Chasing artifacts ᐧ Pulling people away from real work. In this case, what surprised everyone was not the $754K. It was how SPRX surfaced the credits. We gathered and analyzed the technical artifacts the team already had. Plans, models, testing logs, and system development records were pulled straight from their systems. Then, we aligned the work with IRS standards, validated it with an expert, and delivered a defensible report in under a month. ⛔️ No reconstructions. ⛔️ No sampling. ⛔️ No hours-long interviews. If the credit has ever felt too heavy to pursue, there might be meaningful dollars sitting in plain sight. If you want to see what this looked like, I am glad to share more. Send me a DM and we can walk through it together.

Get the latest jobs and industry news

Three really expensive words: I don't qualify. I hear this every single day. In fact, I just talked to a fab shop that does exactly what my other clients do. Three years ago, his CPA told him he shouldn't take the credit. WHY? He wasn't in a position to use credits and it was good guidance then. However, it's not relevant anymore. The key difference between his shop and the next is that one is taking advantage of the R&D tax credit and getting $187,281. Each year stands on its own. Which is why I am always curious when someone says they don't qualify. Often it's because the year they explored it, they weren't paying tax. Stay curious. Ask why.

Two announcements today: A $31M Series B. And a new name: Onshore. SPRX is now Onshore. The name is deliberate. Tax shouldn't feel abstract or outsourced. It should be accountable and grounded in evidence. Here's why we're building this: More than $200B in tax incentives go unclaimed every year. A manufacturer in Ohio leaves millions on the table because their CPA firm stitches together data by hand. An engineering firm in Texas skips R&D credits because the study takes too long to be worth it. The industry rewards time spent, not outcomes delivered. We set out to replace that model entirely. Onshore is the AI-native platform for tax. Not a tech-enabled service, not a better spreadsheet, not more accountants. 500+ companies have trusted us to identify and defend over $600M+ in incentives. Results that speak for themselves. This Series B, led by FPV Ventures with Vertex Ventures US, Restive Ventures, Pruven Capital, Counterpart Ventures, NextEquity Partners and Y Combinator, lets us keep rebuilding the foundation of corporate tax. Starting with credits and incentives, expanding into every workflow the industry still does by hand. Grateful to Wesley Chan, Pegah Ebrahimi & the FPV team; Sandeep Bhadra & Vertex Ventures; Ryan Falvey & Restive Ventures; Jared Moberg, Travis Skelly & the PruVen team; Patrick Eggen & the Counterpart team; Alok Pandey & the NextEquity team; Tom Blomfield & our partners at YC. We're just getting started. Read more in the comments!