AI-Powered Formulation Intelligence

From Empirical Formulation
to Intelligence-Driven
Biopharmaceutical Formulation

FormulytiX is an AI-powered formulation intelligence platform for biologics, vaccines, and complex molecules — combining real-time scientific literature intelligence, a controlled Swiss-hosted AI layer, in silico stability risk modelling, and the Analogous Molecule Engine (AME). All outputs require expert scientific review and are not intended as validated regulatory submissions. Currently in active development.

Explore Modules → How It Works
LEXIS · Excipient Intelligence
mAb Formulation Analysis
mAb SC Aggregation Risk: Low
Representative Output · Conceptual
High
Stability Confidence
ICH Q5C
Guideline Framework
RAG
Cited Sources
Early
Decision Support
🔬
RAG-Based Auditability
🇨🇭
Swiss-hosted AI · Azure CH-North
📋
CTD/IMPD Workflow Support
🛡️
GxP-Aware Architecture
🔗
Native BioXion Bridge
Market Convergence

Why Now Is the Moment
for Formulation Intelligence

Four converging forces create an unprecedented market window for a focused, regulatory-aware formulation intelligence platform.

01
RAG Architecture Maturity
Retrieval-Augmented Generation now delivers traceable, citable outputs — a prerequisite for GxP-regulated pharmaceutical environments where black-box AI cannot be tolerated.
02
Exploding Knowledge Base
Rapid growth of publicly available peer-reviewed literature, patent databases, EPARs, and FDA product review documents creates the infrastructure for scientifically meaningful AI depth.
03
Rising Biologics Complexity
Pipelines shift toward high-concentration SC injectables, bispecifics, and novel modalities — empirical formulation approaches are reaching their practical limits.
04
Phase 3 Attrition Pressure
Industry-wide pressure to reduce late-stage clinical failures driven by stability and manufacturability issues that could have been identified earlier with better decision support.
Eight Specialized Modules

Full-Spectrum
Formulation Intelligence

Biologics, vaccines, and complex molecules — operate modules independently or as a connected intelligence layer from early candidate selection through regulatory submission.

Module A
LEXIS
Literature & Prior-Art Intelligence
Real-time integration with PubMed, EMA, FDA, patent literature, and regulatory guidance repositories. Every output is fully citable, traceable, and grounded in the latest available science — no static corpus to go out of date. Prior-art and patent signals are provided for scientific and formulation awareness only — not freedom-to-operate, patentability or legal advice.
Sequence-to-formulation guidance for mAbs & proteins
Excipient compatibility with confidence-weighted rationale
FDA Purple Book / EMA EPAR regulatory precedent mining
Module B
ROUTEX
Delivery Route & Administration Optimization
Scenario-based modeling of delivery route feasibility. Supports early go/no-go decisions before costly clinical studies — essential for patient-centric SC injectable design.
Viscosity modeling & injectability assessment
Auto-injector and device platform compatibility mapping
High-concentration protein concentration optimization
Module C
STABILIS
Stability Simulation & Risk Modeling
In silico analysis of degradation pathways and long-term stability risk. Operates as a digital decision-support layer alongside ICH stability studies.
ICH Q1A & Q5C study design framework generation
Arrhenius kinetics long-term stability projection
OOS/OOT early signal detection
Module D
DEVIX
Drug–Excipient & Delivery Intelligence
Structured drug–excipient compatibility and delivery-feasibility analysis — excipient compatibility, SubQ viscosity and injection-volume constraints, container-closure and adsorption risk. Does not replace laboratory compatibility studies.
Drug–excipient compatibility matrix with confidence-weighted rationale
SubQ viscosity, injection volume & container-closure constraints
Excipient adsorption & interaction risk flagging
Module E
COMPEX
Comparability Support
Structured comparability reasoning across formulation changes — reformulation, tech transfer, excipient/container change and scale-up. Supports the comparability rationale; does not replace formal comparability studies.
Change scoping & impacted-CQA identification
Reference & prior-formulation evidence mapping
Comparability summary & rationale (expert-reviewed)
Module F
REGIX
Regulatory Mapping & Gap Identification
AI-assisted mapping of formulation outputs to ICH / FDA / EMA / Swissmedic expectations, with gap identification — not regulatory drafting, not legal or regulatory advice. Designed for integration with BioXion for connected formulation-to-CMC workflows.
CTD Section 3.2.P expectation mapping & gap report (EMA / FDA / Swissmedic)
Dossier cross-reference consistency checker
ICH / EMA / FDA guideline currency monitoring
Cross-Cutting Capability

The Analogous
Molecule Engine

A horizontal AI capability operating across all modules — shared molecular intelligence that amplifies every individual module's value. Active across all modules from launch.

Upon submission of a new molecule, the AME identifies the most structurally and therapeutically analogous molecules, retrieves their formulation data, stability profiles, and regulatory intelligence, and pre-populates this as contextual input across all active modules.

⚡ Designed to significantly reduce early-stage formulation intelligence gathering time — evidence-anchored from day one.
LEXIS
REGIX
STABILIS
DEVIX
COMPEX
ROUTEX
AME
Core
Technical Foundation

Designed for Regulatory-Aligned
Scientific Environments

🧠
RAG Architecture
All outputs are designed to be grounded in verifiable, citable scientific literature. Unlike black-box AI, RAG-based responses are intended to be fully traceable to source documents — a key requirement for GxP-regulated environments.
🔬
Domain-Specific LLMs
FormulytiX runs a controlled, domain-specific AI layer — a deterministic rule engine plus retrieval over a curated formulation corpus, with entity masking and model routing, deployed on Azure Switzerland North. A managed open model is used by default, with selective, no-training / no-retention use of a Claude-class model.
🏗️
Modular API-First
Each module exposes a documented open API. AI Drug Discovery platforms can pass candidate molecules directly to FormulytiX, triggering full formulation intelligence workflows — with results flowing onward to BioXion for CMC program execution.
🇨🇭
Swiss Data Residency
Hosted exclusively on Azure Switzerland North. A controlled AI layer with entity masking, model routing and no-training / no-retention terms; a dedicated, fully sovereign Swiss deployment is available as an enterprise option. Data residency within Swiss/EU jurisdiction.
⚖️
Hybrid AI Architecture
Deterministic rule engine (encoding ICH guidelines and stability science) combined with a routed language model (managed open model by default, selective Claude-class) and real-time RAG retrieval. Every output carries confidence tier, source citations, and mandatory expert review flag — fully auditable.
🔗
BioXion Integration
Structured outputs designed to flow directly into BioXion's regulatory dossier module (CTD/IMPD), reducing manual re-entry and cross-document inconsistency risk.
Target Audience

Built for Three
Critical Segments

01
Biotech & Virtual Pharma
Limited CMC Capacity
Organisations with promising biologics, vaccine, or complex molecule candidates but limited internal CMC formulation expertise. FormulytiX is designed to deliver AI-assisted formulation guidance, intended to support early development stages where formulation decisions carry the highest downstream impact.
02
CDMOs
Multi-Project Pipeline Management
Organisations managing formulation development across multiple client programs — biologics, vaccines, and complex molecules — simultaneously. FormulytiX is designed to deliver scalable intelligence intended to support throughput and traceability — a meaningful differentiator where scientific rigor and auditability are valued.
03
Mid-Size Biopharma
AI-Augmented Internal Teams
Organisations with internal CMC teams seeking to reduce Phase 3 stability-related failures, accelerate regulatory timelines, and systematise formulation decision-making across their full pipeline — biologics, vaccines, and complex molecules.
HTS Digital Thread

FormulytiX + BioXion
A Connected Intelligence Layer

FormulytiX
Formulation Intelligence
⟶
BioXion ↗
CMC & Regulatory Execution
CTD 3.2.P Draft
⟶
IMPD Dossier
⟶
MAA Submission

Designed to support CTD/IMPD content development workflows — structured outputs aligned with dossier preparation needs.

🔁
REGIX → BioXion Bridge
REGIX module outputs are structured to flow into BioXion's CTD/IMPD Regulatory Intelligence module — an AI-powered development intelligence platform covering program management, CDMO oversight, quality events, and regulatory dossier preparation. The bridge is designed to reduce manual transcription and cross-document inconsistency.
🔄
Drug Discovery → FormulytiX → BioXion
AI Drug Discovery platforms can pass candidate molecules directly to FormulytiX via open API — triggering full formulation intelligence workflows without manual re-entry. Formulation outputs then flow into BioXion ↗ for CMC program execution. One seamless digital thread from molecule discovery to regulatory submission.
🏢
Distinct, Complementary Roles
FormulytiX focuses exclusively on formulation intelligence, while BioXion ↗ provides broader CMC program intelligence — covering Tech Transfer, Analytical Intelligence, Quality Events, and Regulatory Execution. Both platforms are designed to operate together as a connected intelligence layer, sharing structured outputs across the biologics development lifecycle.
Development Roadmap

Phased Deployment
for Maximum Impact

Phase 1
Biologics & Vaccines Core
  • Monoclonal antibodies (mAbs) — primary focus
  • Recombinant proteins (enzymes, Fc-fusion constructs)
  • LEXIS — Literature & Prior-Art Intelligence
  • STABILIS — Stability Simulation & Risk Modeling
  • Analogous Molecule Engine (AME) core
  • AI Intelligence Dashboard — real-time signal aggregation
  • Validation Module — wet-lab protocol generation & BioXion data export
  • Pilot with 2–3 biotech or CDMO partners
Phase 2
Full Platform Expansion
  • BsAbs, ADCs, viral vectors, mRNA/LNP systems
  • Oligonucleotides, peptides, vaccines, selected small molecules (BCS Class II/IV)
  • ROUTEX — Delivery Route Optimization
  • DEVIX — Drug–Excipient & Delivery Intelligence
  • COMPEX — Comparability Support
  • REGIX — Regulatory Mapping & Gap Identification
  • Open API — AI Drug Discovery platform integration
  • Full BioXion integration & commercial go-to-market
The Team

The minds behind
FormulytiX

A founding team uniting decades of biopharma formulation expertise, computational chemistry, and operational discipline — building the AI-powered formulation intelligence we wished existed.

Sinisa Popov — Founder & CEO
Sinisa Popov
Founder & CEO
15+ yrs biopharma industry experience
External Quality at MSD (Merck & Co.)
PhD Biologics & Vaccines Formulations (active)
MSc Chemistry · Six Sigma Black Belt
Slim Chaker — COO
Slim Chaker
COO · Project Management & Operations
MSc in Engineering
R&D, production & process engineering
CAD & FEM engineering (Creo, Solidworks, Ansys)

Ready to Transform
Your Formulation Process?

Join the pilot cohort of biotech, CDMO, and biopharma partners shaping the future of AI-driven formulation intelligence — across biologics, vaccines, and complex molecules.

Express Your Interest → Learn More at HTS

Important Notice: FormulytiX is designed as an AI-powered formulation intelligence platform for biologics, vaccines, and complex molecules — currently in active development. All outputs require expert scientific evaluation and are not intended for direct regulatory submission. Platform capabilities described represent the intended product scope — not a deployed production system.