# Priya Natarajan, Co-founder & CEO at Fieldpath — read of contract-negotiation-ai, May 18 2026

> 11 years in B2B software, previously PM at Salesforce, second company now. Raising a Series A in Q3. Have twin 4-year-olds and a Notion folder called "dumb things I almost signed."

## How I got here

Someone in a founder Slack I'm in (Operators & Allies, ~800 people) posted a message that just said "has anyone used this for their Series A docs?" with a link and no other context. No replies yet. That's actually how I find half the tools I end up using, so I clicked. I was on my lunch break and had maybe eight minutes.

## What I clicked first

The live result panel stopped me. Not the hero. The hero says "Stop leaving 3 to 5 percent equity on the table every round" and I rolled my eyes a little because that number is doing a lot of work without any source behind it. But then the demo shows a SAFE with a specific pro rata threshold flagged: "Pro rata threshold set at $250k. Market for this stage is $100k." That's an actual, checkable claim. I've seen pro rata thresholds in the wild and $250k is indeed on the high end for seed. That one example did more for me than the entire hero paragraph.

## Where I paused

The four-step flow at "From paste to push-back in under five minutes." Step 3 says "you get drafted push-back language plus the comparable deals that justify it." I stopped there because that is the actual hard thing. Anyone can flag a clause. The question is whether the counter-language is legally defensible or if it's going to make me look like I ran my term sheet through ChatGPT. The page says the language is "written in plain operator voice, not law-firm boilerplate, so it reads like a founder pushed back, not a lawyer." I want to believe that. I don't yet.

## What I distrusted

Two things.

First, the "3 to 5 percent" claim appears three times on the page with zero backing. No study, no dataset, no "we analyzed X term sheets." It's just asserted repeatedly, which is the pattern of a number someone made up that sounded plausible and then kept.

Second, and this one actually stopped me cold: "Honest disclosure: we don't have live customers on this idea yet. We shipped the strategy package; you ship the customer conversations." I had to re-read that sentence twice. So this isn't a product people are using? It's a strategy package for someone to build the product? Then the stuff above, the demo, the "try it" button, the "ask for a refund if it doesn't surface two negotiable points," all of that is describing a product that may not exist in working form? I got genuinely confused about what I was looking at. The page switches midway from talking to me as a founder who needs help on my term sheet, to talking to some other person about whether this business idea has a 71/100 adoptability score and a "Year-1 take-home (Fermi)" of negative $23,800. Those are two completely different audiences and the page is trying to serve both of them and ends up confusing both.

## What would convince me

A recording of an actual term sheet negotiation where the tool's suggested counter was used and the founder can say whether it landed or got laughed out of the room. Not a testimonial. An actual story with enough detail that I can feel whether the counter-language is founder-credible in a room with Sequoia associates. Even one anonymous founder saying "I used the suggested language on my pro rata clause and here's how the conversation went" would move me considerably. Also, if the tool is real and working, show me a non-demo PDF being analyzed live, not a pre-loaded example.

## What I'd ask in an email reply

1. The page says counter-language is written "in plain operator voice." Can you show me two examples side by side, the boilerplate version and the version the tool produces, for the same clause? I want to see whether the output actually sounds like a founder or whether it sounds like a founder who had AI write their email.

2. "Built on patterns from publicly-documented term sheets" -- how recent is the dataset? NVCA forms are years old and market has moved. If I'm comparing my pro rata clause to 2019 market norms I'm negotiating against stale data.

3. The page has a disclosure that there are no live customers yet. Does that mean the analysis engine is live and I can upload my actual term sheet today, or am I looking at a concept that isn't in production?

## Verdict: on-the-fence

The demo example is the best thing on the page, it's specific enough to be useful. But I cannot figure out whether this is a working product or a product idea being sold to operators, and if I cannot figure that out in eight minutes I am not going to hand it my term sheet before a Series A call.

---
*Memo by skeptic persona, generated 2026-05-18. Studio breaks own self-grading loop.*