DEEPC, the AI co-scientist for pharmaceutical formulation
Design, reverse‑engineer and reformulate, with the IP in mind.
Start from a novel API, a novel biologic or a product on the market. DeepC designs the formulation, reconstructs the reference product, moves it to a new dosage form and maps the patents so you can design around them.
A real run
Glucophage 500 mg, reverse-engineered: 317 sources read, 34 cited, 7 minutes 41 seconds from question to answer.
Glucophage 500 mg deformulation
Partners and programmes
A study for every question your scientists ask.
A quick excipient check comes straight back; a formulation, an FTO opinion or a safety report is run as the study built for that work. Either way the answer comes back cited.
From spray drying to subcutaneous mAbs.
DeepC works across oral solids, sterile injectables, inhaled, topical and ophthalmic products and biologics, and the processes that make each of them.
- Spray drying
- Hot-melt extrusion
- Wet granulation
- Roller compaction
- Direct compression
- Film coating
- Capsule filling
- Lyophilisation
- Monoclonal antibodies
- Antibody-drug conjugates
- mRNA-LNP
- AAV gene therapy
- Injectables
- Prefilled syringes
- Subcutaneous autoinjectors
- Infusions
- Inhalation
- Nasal sprays
- Transdermal patches
- Topicals
- Ophthalmics
- Softgels
- Oral suspensions
- Vaginal rings
Where every answer comes from.
Three ways to check an answer
Superscripts in the answer anchor to a Sources list beneath it, split into Cited and Consulted and grouped by type. When several records support one claim, each is cited on its own, so each can be checked.
When a figure in a regulatory document settles the question, DeepC shows it in the answer with its page and source. It shows a figure only after it has opened and read the document.
A citation to one of your own documents opens it in the document viewer at the section the answer drew from, highlighted.
A library built for formulation.
Regulations from medicines agencies around the world, patents, toxicology, excipient and pharmacopoeial standards, physicochemical data, clinical evidence and your own documents, in one library built by DeepCeutix and cited in every answer.
Your next experiment, proposed.
DeepC Optimize turns a formulation into an optimisation protocol. You make and test the batches; it proposes each next round from everything measured so far, inside limits screened against regulatory precedent.
Proposals inside the formulation limits
49%
Unconstrained optimiser
Searches the full design space.
100%
DeepC
Searches inside the formulation limits.
We know your IP is everything.
So is keeping it yours.
EU data residency is available, and the AI model providers behind DeepC keep none of your research.
- EU data residency available
- Organisations that need it can have their content hosted and stored in the European Union.
- Zero data retention
- Third-party AI model providers use your content only to return a response. They retain none of it and train none of their models on it.
- An isolated sandbox
- Data analysis runs in an isolated sandbox, one per conversation, with no network access.
- Run from your Microsoft directory
- Entra ID, formerly Azure Active Directory, adds, changes and removes accounts, and signs your team in with their work accounts.
- SharePoint permissions, mirrored
- Each person sees in DeepC the documents they can open in SharePoint.
- A second factor on every account
- A passkey or an authenticator app from the first sign-in.
Built for regulated research.
- HIPAABAA available
- GDPR & UK GDPRDPA available
- CCPA / CPRACalifornia consumer rights
- FDA 21 CFR Part 11Electronic records & signatures
- GxP-aligned SDLCGAMP 5 risk-based validation
Book a demo with a formulation lead.
From a first call to a pilot on your own programme.
01A demo
A call with a formulation lead, who is in touch within a business day.
02A pilot
On your own programme, scoped with your team.
The AI model providers behind DeepC retain none of your content and train nothing on it.
Prefer email? info@deepceutix.com