In this episode of The Sunny Ray Show, host Sunny Ray talks with Nigel Blair, founder of Jian AI, an artificial intelligence platform built for the full life sciences R&D journey, from early research idea through postmarket surveillance. Blair traces the company's origin to his own cancer diagnosis and a decade working in medicine, where he saw roughly 90% of drug developers fail to bring products to market. Jian AI aims to raise those odds using guardrailed, agentic AI combined with a CRM, while guaranteeing no LLM knowledge leakage so researchers' intellectual property never trains outside models like those from Anthropic or OpenAI. Blair describes pursuing ISO 42001 certification and a SOC 2 audit before generating revenue, a deliberate bet on compliance to win institutional trust. He walks through the platform's rocky first iteration, lessons learned from thirty pilot life science companies, and an upcoming relaunch featuring project mind mapping. The conversation also covers Jian AI's admission to the NVIDIA Inception program and an AWS Activate partnership exploring global co-selling, plus Blair's vision for AI accelerating medical breakthroughs within five years.
A cancer survivor turned founder builds guardrailed AI to fix life sciences' 90% drug development failure rate.
What is Jian AI's elevator pitch, and why does it matter?
Years ago I went through a medical process that started with cancer, and it pushed me toward doing something positive. I had already worked ten years in the medical field and saw that about 90% of companies developing life saving drugs were failing. So we set out to use artificial intelligence to improve researchers' chances of success across the R&D journey, from idea through postmarket surveillance. It's really a societal mission.
Life sciences R&D has been digitized for two decades. What is still done the slow way that shouldn't be?
Life sciences is complex. Initial research can take years or decades, then you face ethics committees and clinical trials, then regulatory approval. A medical device can take five to seven years to reach market, and a drug can take up to fifteen. It's a fragmented, siloed system with many steps. We designed our platform to reduce that complexity and provide guidance across the entire journey, from research through postmarket surveillance.
Your pitch rests on no LLM knowledge leakage. What does that actually mean to a researcher?
Researchers everywhere, in the US, Canada or Australia, are developing new intellectual property, and many feed those ideas into large language models from companies like Anthropic or OpenAI. We built our platform to be secure, going through a full ISO process, so when researchers use our site their data does not train those large language models.
What was the first version of Jian AI, and how wrong was it?
The first version was me sketching on paper, and honestly it was pretty wrong at the start. We vibe coded a beta, then spent five to six months working with about thirty life science companies, big institutions and small researchers, refining the product. We did a soft launch last June, but I wasn't happy with that first iteration, so this week we're launching a version I'm much happier with.
You got ISO 42001 certified before you had revenue. Why do that in that order?
In life sciences you deal with a lot of technical data, and institutional clients expect certain standards before they'll work with you. We focused on building the foundation first, so we have ISO 42001, we've been audited for SOC 2, and we're being audited on AI compliance standards in September. That shows customers we take compliance and process seriously.
What's the hardest engineering problem still unsolved for you?
Over the next two to four weeks, in two week sprints, we're building a system where you load a template anywhere in the world and it populates against your own data plus the health and regulatory data for that country. So if I'm in South Africa selling into America, the system gives me a first cut on documents pointed to the right regulations, though we always keep a human in the loop.
What changes on day one for a team that switches to Jian AI?
You get to play with the system and it mind maps whatever you're working on, holding your documents as a source of truth while you control who is invited in. It also lets you load and populate template documents. Really it reduces your time on administration, reduces cost, and gives you new knowledge you didn't have before, whether you're a single researcher or a VC watching a whole portfolio.
If Jian works, what does a life sciences team look like in five years that it doesn't today?
Right now about ten thousand papers a day flood the market, and it's impossible for people to review them all. In five years you'll get prompts from our AI agent, we call it PIP, telling you what to look for based on your own project. That will speed up completing research and lead to a lot more advances in medicine and new products.
Building something daring? Sunny talks to founders like this every day. Fifteen minutes to see if your story belongs on the stage.
Claim your pre-interview