Answers from the documentation; checking which model is on duty
Hi - I answer from the OpenSmartRoute documentation: routing, the API, plans and quotas, self-hosting. Ask away, or open a support ticket if you need a person.
Grounded in the docs - follow a source before acting on it.
SAS and Posit link tools for auditable AI in clinical research - OpenSmartRoute
Introduction - announcement of SAS and Posit partnership for clinical research
SAS and Posit have announced a new partnership to improve how clinical research data is handled and analyzed. The goal is to make AI tools more transparent and easier to audit. This partnership connects SAS’s clinical research technology with Posit’s open-source development tools. It aims to support regulated workflows that combine SAS and R programming languages. The partnership was announced on November 16, 2026.
This collaboration brings together SAS’s expertise in analytics and data controls with Posit’s open-source environment for mixed-language workflows. It is designed to help teams working in regulated clinical environments keep control over their data and code. The partnership also aims to improve reproducibility and compliance in clinical research. It is part of a broader effort to make AI-assisted workflows more transparent and trustworthy in regulated settings.
Integration details - connecting SAS Viya and Posit Workbench for mixed-language workflows
The integration connects SAS Viya, a cloud-based analytics platform, with Posit Workbench, an open-source development environment. This setup allows users to write code in both SAS and R languages within the same environment. The connection is made through tools like Positron, Posit Connect, and Posit Package Manager.
SAS contributes its analytics capabilities, access controls, and audit features. Posit provides the open-source environment where programmers can develop and run mixed-language code. This setup enables seamless collaboration between SAS and R, along with Python and SQL. It allows teams to work in a unified environment while maintaining control over their code and data.
The integration also supports versioned environments for R and Python. Teams can develop code with Positron and Posit Assistant, which include AI-assisted coding tools. This helps reduce errors and speeds up development. The environment ensures that dependencies and interoperability are managed consistently, making workflows more reliable and reproducible.
Shared execution and environments - how SAS and R code run together securely
Saturday Night Live aired a sketch featuring impressions of major tech executives, humorously exploring spycraft and industry scandals.
In this integrated setup, SAS Viya executes SAS code under the permissions of the programmer working in Posit Workbench. The programmer can write SAS and R code in the same session, with Viya running the SAS code securely. This means that the code runs in a controlled environment, respecting user permissions.
The connected environment also supports running Python and SQL code alongside SAS and R. This flexibility allows teams to choose the best language for each task. The environments are managed to keep dependencies consistent, which helps ensure that results are reproducible. Teams can develop and test code in a managed setup before deploying it for production or submission.
This shared execution model helps avoid the need for workarounds or copying data between different systems. It also reduces the risk of errors and security issues. Programmers can focus on their analysis, knowing that their code runs securely and within governed environments. This setup supports compliance with regulatory standards by keeping all actions traceable and auditable.
AI-assisted data standardization - workflows for data mapping and code generation
AI tools are integrated into the data standardization process to help prepare research data for regulatory submission. These tools assist in mapping raw data to standardized formats, reusing transformation assets, and generating SAS code for data processing. The goal is to make data preparation faster and more consistent.
The workflow begins with AI recommending data mappings based on existing standards. It suggests how raw data should be transformed to meet regulatory requirements. The AI also helps reuse previous transformation assets, saving time and reducing errors. Once the mappings are approved, the system generates reviewable SAS code that can be inspected and modified by human experts.
Human reviewers remain in control of the process. They review the AI-generated code and decisions, ensuring that the output meets quality and compliance standards. The process creates a traceable record of decisions, which is essential for audits. This approach combines AI efficiency with human oversight, making data standardization more reliable.
Data controls and compliance - access, signatures, audit trails for regulatory standards
Managing access and ensuring compliance are critical in regulated clinical research. The integrated workflows include role-based permissions, electronic signatures, and detailed audit trails. These controls help teams meet strict regulatory standards, such as those from the FDA.
Access controls specify who can view or modify data and code. Electronic signatures provide a secure way to approve changes or data outputs, with traceability for audits. Version control tracks modifications over time, making it easy to review the history of a project. Audit trails record every action taken within the system, including data access, code execution, and approvals.
Viya’s governance features support compliance with regulations for electronic records and signatures. The system ensures that all actions are documented and can be reviewed during audits. These controls help teams maintain data integrity and transparency, which are essential for regulatory approval and data sharing.
Regeneron’s experience - real-world benefits of integrated workflows
Regeneron, a major pharmaceutical company, has used these integrated workflows to improve their clinical programming. Previously, they relied on workarounds to manage mixed-language workflows involving R and SAS. These workarounds often involved copying data and managing multiple access controls, which added complexity and risk.
With the new integrated environment, Regeneron programmers can write both SAS and R code in the same session. They do so under their own access rights, with AI assistance operating within those controls. This setup reduces overhead and simplifies workflows. It also helps ensure that all actions are traceable and compliant with regulations.
Dilip Raghunathan, a business solutions partner at Regeneron, noted that the new system allows for seamless handling of complex multilingual workflows. It removes the need for manual workarounds and allows scientists to focus on their research questions. The result is more efficient programming and better compliance with regulatory standards.
Future plans - expanding support for clinical research workflows
SAS and Posit are continuing to work with their joint customers to expand the range of supported workflows. The immediate focus is on building a shared framework that supports language choice, AI assistance, access control, and reproducibility. This framework aims to cover more aspects of regulated clinical research.
The companies plan to support additional workflows beyond the initial integration. This could include more data management processes, analysis steps, and reporting functions. The goal is to create a comprehensive environment that meets the needs of regulated clinical research teams.
The partnership’s wider significance depends on how many workflows can be brought into this shared framework. By doing so, SAS and Posit aim to make regulated clinical research more transparent, efficient, and compliant. They seek to enable teams to adopt open-source tools without sacrificing control or auditability.
How OpenSmartRoute helps - routing requests for secure, compliant AI processing
OpenSmartRoute is an open-source router designed for AI requests. It manages how requests are sent to different models, agents, or tools. The router scores each candidate based on quality, cost, speed, and safety. Teams set the importance of each factor for every request.
Using OpenSmartRoute, teams can ensure that personal data stays on-premises or within a region. The system enforces rules to prevent crossing cost caps or region boundaries. It learns from outcomes, so models that perform well get more traffic. New models can be added easily, competing on the same requests.
The hosted platform keeps a catalog of models, prices, and rankings. It also provides tools for measuring routing accuracy and spotting prompt injection or personal data leaks. This setup helps teams run AI workflows securely, with compliance and cost control built in.
How it compares - what existed before, what this changes and what stays the same
Before this partnership, SAS and Posit offered separate tools for clinical research and data analysis. SAS provided a governed environment with strict access controls, audit trails, and compliance features. Posit (formerly RStudio) focused on open-source development environments for R, Python, and SQL. Teams often had to create workarounds to combine SAS and R workflows.
The new integration links SAS Viya and SAS Clinical Acceleration with Posit Workbench and related tools. This creates a shared environment where SAS and R code can run together seamlessly. It allows programmers to write in both languages in one session, under their own permissions. AI-assisted data standardization and transformation tools are now accessible within this unified setup.
What stays the same is the focus on compliance, traceability, and control. SAS’s governance features remain central, ensuring that all data handling meets regulatory standards. Posit’s open-source environment complements this by enabling flexible, reproducible workflows. The partnership does not replace existing tools but enhances their interoperability.
This integration changes how teams work with SAS and R. It reduces the need for manual data transfers, duplicate access controls, and complex workarounds. Programmers can now develop and review code in a single environment, with AI assistance operating within their permissions. It also broadens the scope of workflows that can be managed under compliance rules.
Questions this leaves open - what the source does not say and how a reader can check it
The source does not specify the exact technical details of how SAS Viya and Posit Workbench are connected. It does not describe the underlying architecture or whether specific APIs or middleware are used. Readers interested in implementation should consult technical documentation from SAS and Posit.
It also does not mention the limits of the current integration. For example, it is unclear how many languages or tools can be used simultaneously, or what the performance implications are. Teams should test the setup in their environment to see how it handles complex workflows or large datasets.
The source does not specify the licensing or cost implications of this integration. It is not clear whether additional licenses are needed or if the integration is included in existing subscriptions. Organizations should check with SAS and Posit sales or support teams to clarify licensing terms.
Another open question is how AI-assisted standardization workflows are evaluated. The source mentions recommendations and code generation, but does not detail accuracy, review processes, or validation steps. Teams should establish their own testing and review procedures to ensure AI outputs meet quality standards.
Finally, the source does not describe how this integrated environment scales for larger teams or more complex projects. It is important to verify whether the setup supports multiple users, version control, and collaboration at scale. Pilot testing can help determine if the system fits the organization’s size and needs.
To explore these points, teams can review technical documentation, conduct pilot projects, and engage with vendor support. This helps ensure the integration aligns with their regulatory, technical, and operational requirements.
What to do - steps for teams to adopt these integrated tools
Teams interested in adopting these integrated workflows should start by evaluating their current data and coding practices. They can explore connecting SAS Viya with Posit Workbench in their environment. Setting up shared environments and access controls is a key first step.
Next, teams should consider implementing AI-assisted data standardization workflows. Training staff on how to review and approve AI-generated code ensures compliance. They should also establish audit trails and signature processes to meet regulatory standards.
Finally, integrating request routing with OpenSmartRoute can further enhance security and cost management. Teams can set rules for data handling and model selection, ensuring that AI processing remains compliant. Starting with small pilot projects helps test and refine the setup before scaling.
Claude Fable 5.2 creates images by simulating brushstrokes through Python code, mimicking physical painting. It can produce complex, photorealistic art and plan physical tasks.