Skip to main content

Intelligent infrastructure for life sciences

Between the brilliance of discoveryand the precision of a regulatory submission.

Coralis Bio is the bridge. We’re building cloud-native, agentic AI-assisted software for the scientists, engineers, and regulatory teams developing advanced therapies from discovery toward the clinic.

Laboratory equipment in a controlled clean-room environment
What we believe

There are hundreds of decisions standing between advanced therapy and the patients who need it. We’re building software to help the people behind them and make the process more organized.

The science going in is unpredictable, brilliant, prone to failure. The output has to be a drug product that is safe, effective, and identical for every patient, regardless of the scale at which it’s manufactured. Coralis is the structured space in between — honest about what it knows, careful about what it doesn’t, and built as a peer to the people who put their names on the science.

CMC
Process development, analytics, and manufacturing readiness.
Nonclinical
Safety, pharmacology, and translational study design.
Clinical
First-in-human and pivotal study strategy for advanced therapies.
Regulatory
FDA, EMA, and Health Canada engagement, INTERACTs through INDs.
Orientation

What Coralis is building.

Who we are

A small team of CGT development leaders, engineers, and designers with decades of combined experience across FDA, EMA, and Health Canada programs. We’ve worked inside the kinds of companies we’re now building for.

What we’re building

Cloud-native, AI-assisted software for advanced therapy teams — built as a peer to the people who already do the work, helping small CGT companies make sense of their own programs with a clear sense of what software should and shouldn’t decide.

How we work alongside you

A lot of what matters in regulatory work and documentation isn’t published anywhere — it lives with practitioners, in judgment calls and precedent learned inside real programs. Our team works with yours, not just the software.

Where we’re going

Private beta with a small group of design partners through 2026. We’re building deliberately, in close collaboration with the people who will use what we make.

How we think about AI

Any AI can tell you what to do.The sequence is the work.

Any chatbot can draft a regulatory roadmap in sixty seconds. The output looks correct, reads well, and is routinely wrong on the details that matter. In cell and gene therapy, the draft isn’t the hard part — the sequence is, and so is the evidence behind each step.

Coralis is built around a structured record of what a program needs to do, and in what order. AI drafts inside that structure. Every output carries the model that produced it, its confidence, and its sources. A reviewer can accept, correct, or reject it — and the record remembers which.

  1. Track the critical path

    Milestones, dependencies, and the work required between them, versioned by process.

  2. See the work

    One view across clinical, nonclinical, CMC, and regulatory, current with every update to the record.

  3. Ask the record

    Query precedents, filings, and your own program. Answers cite sources and a confidence level.

  4. Draft the filings

    INTERACTs, pre-INDs, INDs, tech transfer plans. Generated from the record, signed off by your team.

Built on private infrastructure

Private AI. Isolated data. Your keys.

Coralis is built for organizations that cannot afford to compromise on data security. Client isolation, encryption, and access control are not features we added — they are structural properties of how the platform works.

Get in touch
  • Private AI, never public

    Your documents are processed by private, enterprise-grade AI models running on isolated cloud infrastructure. Never sent to public AI services. Never used to train models.

  • Per-organization isolation

    Every client gets its own isolated database and its own encrypted storage container, with an encryption key unique to your organization. There is no shared data layer.

  • Single-region, infrastructure-as-code

    All processing and storage occurs within a single secure cloud region. The entire infrastructure is defined in code, version-controlled, and fully auditable.

  • Auditable by design

    AI queries, document processing events, and access patterns are logged for compliance and review. Every output carries its provenance.

The team

Built by operators. More than software.

Coralis is led by Dr. Lara Ionescu Silverman, and Steve Goldenberg, with experience spanning cell and gene therapy development, regulatory strategy, tech transfer, scale-up, software engineering, and agentic AI. Our partners get more than access to the platform — they work directly with the scientists, engineers, and builders behind it.

We are building the software platform we wished existed earlier in our careers. We are working closely with a select group of early partners in a hands-on model shaped by real-world experience.

Scientist working at a laboratory bench
Get in touch

Built with early partners.

We’re working in close collaboration with a small group of early partners through 2026. If you’re a scientist, operator, press, or prospective design partner — we’d like to hear from you.