Unlocking AI Access: How Thomson Reuters Democratized AI with Open Arena

This heading captures the essence of the post, emphasizing both the transformational nature of AI and the specific platform (Open Arena) that was developed to make AI accessible to a broader audience within Thomson Reuters.

Unlocking AI Democracy: The Journey of Thomson Reuters’ Open Arena

This post is cowritten by Laura Skylaki, Vaibhav Goswami, Ramdev Wudali, and Sahar El Khoury from Thomson Reuters.

Thomson Reuters (TR) has long been at the forefront of innovation, providing trusted content and workflow automation solutions for over 150 years. The company recognized early on that artificial intelligence (AI) would fundamentally reshape professional work, as highlighted in TR’s 2025 Future of Professionals Report. A staggering 80% of professionals expect significant changes to their work environments due to AI. With an anticipated 12 hours of productivity gains per week by 2029, TR understood the necessity for a solution that democratizes AI creation across its expansive workforce.

In this post, we’ll explore how TR harnessed the power of Open Arena, a scalable no-code AI solution powered by Amazon Bedrock and other AWS services, to address key business use cases and support various user profiles.

Introducing Open Arena: No-code AI for All

Open Arena was developed to empower non-technical professionals to create customized AI solutions seamlessly. With cutting-edge AI powered by Amazon Bedrock, users can now thrive in a no-code environment, exemplifying TR’s mission to democratize AI access.

Key Features of Open Arena

• High Adoption: Approximately 70% of employees engage with the platform, with a robust 19,000 monthly active users.
• Custom Solutions: Thousands of tailored AI solutions have been created without coding, enhancing both internal workflows and customer products.
• Self-Served Functionality: Users can independently develop, evaluate, and deploy generative AI solutions without requiring technical expertise.

The Open Arena Journey: From Prototype to Enterprise Solution

Initially conceived as a quick prototype, Open Arena was birthed in under six weeks in early 2023 by TR Labs, TR’s dedicated applied research division. The aim was to enable internal exploration of large language models (LLMs) and uncover novel use cases by combining LLM capabilities with TR’s proprietary data.

The launch of Open Arena sparked a noticeable increase in AI awareness and collaboration between developers and Subject Matter Experts (SMEs), resulting in rapid capability development. Responding to demand, TR created an enterprise version of Open Arena built on the TR AI Platform, thereby enhancing security and scalability while maintaining compliance with governance standards.

“The foundation of the AI Platform has always been about empowerment; it’s about enabling users of any background to leverage and create AI solutions.”
– Maria Apazoglou, Head of AI Engineering, CoCounsel

As of July 2025, Open Arena is utilized by 19,000 users each month, reflecting its growing acceptance and utility across the enterprise.

Addressing Key Enterprise AI Challenges

Open Arena has armed thousands of professionals with the tools required to leverage generative AI. The platform effectively tackles four critical enterprise AI challenges:

1. Enablement: Delivers a consistent LLM and service experience, supporting non-technical users across various personas.
2. Security and Quality: Ensures that AI solutions meet governance policies through monitoring and evaluation services.
3. Speed and Reusability: Streamlines workflows, allowing for the reuse of existing AI solutions and prompts.
4. Resource and Cost Management: Tracks generative AI resource consumption, promoting transparency and efficiency.

The platform currently supports numerous AI experiences, including tech support, content creation, and problem-solving, catering to diverse organizational needs.

Why Thomson Reuters Selected AWS for Open Arena

TR strategically chose AWS as a primary cloud provider for several reasons:

• Comprehensive AI/ML Capabilities: Amazon Bedrock simplifies access to high-performing foundation models.
• Enterprise-Grade Security and Governance: Advanced security controls ensure models and data are handled with care.
• Scalable Infrastructure: A serverless architecture offers automatic scaling and pay-per-use pricing to optimize costs.
• Established Relationship and Expertise: A strong partnership between TR and AWS bolsters their collaborative efforts.

“Our partnership with AWS and their innovative services made them the natural choice to power Open Arena.”
– Maria Apazoglou, Head of AI Engineering, CoCounsel

Open Arena Architecture: Scalability, Extensibility, and Security

Designed for a diverse enterprise audience, Open Arena prioritizes scalability and security while maintaining simplicity for non-technical users. The architecture enables enterprise-grade performance with a clear separation between capability and usage.

Key Components of the Solution Architecture:

• No-Code Interface: Intuitive UI with visual workflow builders and drag-and-drop functionality.
• Enterprise Integration: Seamless integration with TR’s existing AI Platform and enhanced security for data handling.
• Solution Management: Features include a searchable repository and usage analytics.

Open Arena leverages various AWS services, including Amazon Bedrock and AWS Lambda, to manage orchestration and user data securely. Each user operates within their own AWS account to comply with TR’s governance policies.

Evolution of Open Arena: From Classic to Amazon Bedrock Flows-Powered Chain Builder

Open Arena has evolved into two versions:

1. Open Arena v1 (Classic): A simple form-based interface for basic prompt customization.
2. Open Arena v2 (Chain Builder): A robust, visual workflow builder that allows users to design complex, multi-step AI workflows.

The integration of Amazon Bedrock Flows as a core feature supports intricate workflows, enabling secure, scalable execution across user accounts.

What’s Next?

TR is committed to expanding Open Arena’s capabilities through a strategic partnership with AWS by:

• Driving further adoption of DIY capabilities.
• Enhancing workflow creation flexibility in the Chain Builder.
• Developing new templates for common tasks.
• Improving collaboration features and extending multimodal capabilities.

“We push the boundaries of what’s possible with Open Arena.”
– Maria Apazoglou, Head of AI Engineering, CoCounsel

Conclusion

Thomson Reuters’ Open Arena exemplifies the successful democratization of AI within an enterprise, leveraging AWS services such as Amazon Bedrock. With 19,000 monthly active users and an impressive 70% employee adoption rate, it illustrates that no-code AI solutions can deliver enterprise-scale impact while maintaining robust security standards.

The innovative architecture and strategic cloud partnerships showcase a blueprint for AI democratization, empowering professionals of all skill levels to harness generative AI in their daily work.

About the Authors

• Laura Skylaki: PhD, leads the Enterprise AI Platform, enhancing product value through GenAI services.
• Vaibhav Goswami: Lead Software Engineer, focuses on democratizing AI access for users across the enterprise.
• Ramdev Wudali: Distinguished Engineer, specializes in enabling users to develop AI solutions.
• Sahar El Khoury: Director of AI Platform Adoption and Training, guides user onboarding and success.

For more insights and updates on Open Arena, stay tuned!