about us
Our Story at Seedless
Most breakthrough technologies fail not because they don't work, but because they can’t be properly tested. In regulated industries (legal, healthcare, finance) the very data needed to evaluate AI tools is locked behind privacy walls. We built Seedless to break through that barrier.
The Problem We Lived
Seedless was born from years of frustration with a problem that shouldn't exist: the inability to evaluate transformative technology because the data needed to test it is locked away.
Our CEO, Josh Kreamer, experienced this bottleneck from every angle. As a lawyer, he watched organizations purchase inferior tools because they couldn't access client data to evaluate alternatives. At AstraZeneca, as Head of Legal Services and Technology Strategy, he watched promising AI deals stall because companies couldn't access data for proper proofs of concept.
Our CTO, Shahrukh Tarapore, encountered the same barrier from the technical side. At Lockheed Martin, he witnessed the staggering cost of creating simulation data for testing, and how insufficient test data crippled sophisticated systems. In healthcare analytics, he watched promising AI models struggle because edge-case testing data simply wasn't available.
We both realized the same truth: data isn’t just a nice-to-have, it’s a critical requirement to innovation and adoption of AI.
See How Simulated Data Works
Josh Kreamer
chief executive officer
a white check mark icon
Harvard Law grad and established innovator in legal tech space
a white check mark icon
Former Head of Legal Services and Technology Strategy at AstraZeneca
a white check mark icon
15+ years solving unstructured data challenges in finance and pharma
Shahrukh Tarapore
chief technology officer
a white check mark icon
Published expert on AI/ML for modeling and simulations
a white check mark icon
Former Lead Software Engineer at Lockheed Martin
a white check mark icon
Serial entrepreneur who bootstrapped a tech startup in healthcare to $8M revenue
Why Simulated Data Changes Everything
Companies are drowning in business communications: emails, chats, contracts, reports. But privacy regulations, compliance requirements, and legitimate security concerns mean that data can't be used for demos, evaluations, or training AI tools.
Real business data isn't isolated. It's interconnected.
The synthetic data industry tried to solve this, but their approach fell short. Traditional synthetic data generation creates isolated artifacts, a contract here, an email there, with no connection between them. This approach doesn’t work for testing AI tools that use context to find the stories hidden in the data. Synthetic data generation also typically requires a seed set of customer’s private data as a starting point.
When an executive sends an email about a merger, that decision ripples through meeting minutes, contract negotiations, compliance reports, and Slack conversations. When a clinical trial participant reports a side effect, it appears in physicians’ notes, lab reports, and adverse event filings. This web of connections is what makes business data valuable for testing AI, and it's exactly what traditional synthetic data can't replicate. of a div block.
our thesis statement
Seedless was founded on a patent-pending breakthrough: using multi-agent AI simulations to generate data with genuine narrative coherence. Our AI agents don't just create documents, they simulate entire business environments where decisions have consequences, relationships evolve, and information flows naturally across communication channels and document types.