AURA - the Business Expert 

Empowers any user by leveraging Enterprise Knowledge & Know-How

LEO - The Engineer

Effectively solve technical challenges in all engineering disciplines

MARIE - The Scientist

Bring Deep Scientific Knowledge to Engineers or Researchers

Challenge
Life sciences scientists are execution-rich and time-poor. Designing and documenting a single experiment can take an hour or more before any real work begins — protocols get written by hand, plans get reformatted, and notes get translated into structured records well after the work is done.

Solution
MARIE turns plain-language descriptions of your work into structured, searchable digital records in real time. Describe an experiment, and get a ready-to-run design, an analysis of the resulting data, and a proposed next step — all traceable to source and logged automatically in a secure, built-in audit trail.

Benefits

• 10x faster authoring and design – Experiment planning drops from ~1 hour to 5 minutes
• More experiments, more learning – Time saved on documentation goes straight into running science
• Faster path to what matters – More data means a quicker route to the candidates and formulations that count
• Audit-ready by default – Every experiment and decision stands up to inspection

Challenge
Life sciences quality teams are data-rich and insight-poor. Quality investigators burn hours rebuilding context before analysis even starts, categorization drifts between analysts, and systemic issues surface too late - as regulatory findings, delayed batch release, or brand damage.

Solution
MARIE turns raw quality data into plain-language answers. Ask a question, get the what, the why, and the next action, synthesized in real time from a single source of truth with data integrity, digital continuity and automated audit trails built in. Every answer traces back to source, so speed never costs defensibility.

Benefits

• 5x faster triage and investigation - Investigators apply judgement instead of reconstructing context
• Earlier systemic detection - patterns visible now, before they become recalls or findings
• Audit-ready by default - every decision stands up to inspection
• Protects what matters - patient safety, treatment efficacy, and brand reputation

Challenge
Regulatory authoring is expert work buried under data handling. Teams spend their time transcribing figures from enterprise systems, chasing source data across silos, and wrestling documents into house style instead of applying the scientific judgement only they can provide. Every manual hop introduces transcription error and breaks the traceability chain that submissions depend on.

Solution
MARIE automatically extracts structured data from enterprise sources and maps it into 3DEXPERIENCE with full traceability. No manual transcription, no broken lineage. It then generates and refines content in the corporate voice, turning unstructured inputs into review-ready output. Authors stop moving data and start providing accurate expert opinion.

Benefits

• Up to 88% faster CMC report authoring - throughput without adding headcount
• Fewer errors - automated data mapping removes manual transcription mistakes
• Full traceability - every value ties back to source, inspection-ready
• Consistent corporate voice - house style applied automatically

Challenge
Configuring molecular modeling studies often requires navigating complex interfaces, customizing molecular visualizations, and selecting the right methods or parameters, taking valuable time away from scientific research.

Solution
A scientifically-aware AI agent understands molecular structures, modeling and simulation methods, and scientific context. For instance, scientists simply describe how they want to view a structure, and MARIE instantly updates molecular visualizations, changing representations, colors, backgrounds, and display styles without manual navigation.

Benefits
Scientists can explore structures more intuitively, spend less time adjusting visualization settings, and focus on interpreting molecular insights. MARIE allows a more intuitive user experience and more time focused on science.

Challenge

Existing Scientific AI assistants can struggle to provide scientifically accurate answers because they rely on inferred information that may not reflect an organization’s unique knowledge and practices. At the same time, scientists must manually identify the right Pipelines, understand how tools and processes are connected, and execute complex data processing and property prediction tasks. Without AI grounded in customer-specific science and computational capabilities, organizations risk slower research progress and reduced productivity.

Solution

MARIE can be equipped with Pipelines as AI tools, next-generation Pipeline Pilot protocols authored in 3DEXPERIENCE Cloud with BIOVIA Pipeline Creator. Through natural language interactions, MARIE can discover and execute these Pipelines using customer-specific methods, processes, and scientific expertise. These Pipelines provide on-demand access to preconfigured data processing, property prediction, and computational analysis capabilities.

Benefits

MARIE delivers answers grounded in scientific truth, powered by an organization’s knowledge, methods, and know-how. By providing AI-driven access to familiar Pipeline Pilot capabilities through customer-specific Skills and Pipelines, MARIE improves scientific accuracy and trust, while democratizing enterprise knowledge and know-how. It simplifies complex computational processes, accelerates the computational modeling lifecycle, and enables researchers to focus more time on advancing discovery.

Challenge
Lead optimization is one of the most time and resource intensive stages of drug discovery. Teams synthesize and test hundreds of molecules through repeated Design-Make-Test-Analyze cycles to improve potency, selectivity, and developability before identifying a viable candidate.

Solution
BIOVIA Generative Therapeutics Design combines generative AI with machine learning–driven multiparameter optimization to design and prioritize molecules before synthesis. Each candidate is optimized against your target product profile, including binding affinity, selectivity, toxicity, pharmacokinetics, synthesizability (retrosynthetic feasibility), and 3D structural requirements. Through connected virtual and real cycles, active learning continuously updates as new experimental data is generated, capturing each organization's scientific knowledge and accelerating Design, Make, Test, Analyze (DMTA) cycles from years to months.

Benefits

• 22%–28% reduction in lead optimization projects: Advance higher-quality candidates to synthesis faster
• Up to 90% fewer compounds synthesized: Focus laboratory resources on molecules most likely to succeed
• Up to 50% faster time-to-market:  Eliminate costly dead ends earlier in discovery
• Iterative design cycles – Active learning improving future predictions based on experimental data

What's New?

• Automatic Scientific Notebook integration: Completed experiments are automatically documented in Scientific Notebook with experiment details, parent references, molecules, and parameters.
• Expanded AI safety profiling: New toxicity and anti-target prediction models, including five CYP models, help optimize target product profiles earlier in the design process.

Challenge
Finding a protein that binds a target is only the first step. Many candidates fail later because of poor stability, specificity, manufacturability, or other developability challenges, increasing costs and extending development timelines.

Solution
BIOVIA Discovery Studio Simulation brings state-of-the-art generative AI and physics-based modeling together in a unified workflow. Scientists can predict structures of protein-ligand complexes with OpenFold3, generate novel binders with RFdiffusion, optimize sequences with ProteinMPNN, and evaluate candidates using protein language models — all within a single environment. Every proposed protein or antibody can be supported by structural insights that help scientists understand why a design is predicted to bind, behave, and develop successfully.

Benefits

• 10x faster time to research-ready structures – Accelerate the path from binder to higher-quality proteins and antibodies
• Drug-like by design – Affinity, specificity, and developability are optimized together, early in the process
• Explainable structural predictions — Understand the structural basis behind every proposed design
• Tackle challenging targets — Generate binders for difficult epitopes beyond the reach of traditional approaches

What's New?

• Predict Immune Structures – A new protocol that generates 3D structural models of antibodies, T-cell receptors, and nanobodies using ABodyBuilder3, TCRBuilder2, and NanobodyBuilder2.
• Analyze Protein-Ligand Complexes – A new protocol that superimposes a set of protein structures, calculates properties, and displays the aligned binding site.