Migrating from OpenAI to AWS creates a powerful platform built on a robust, stable, reliable system.

About Trisk

Trisk is a SaaS company that specializes in no-code AI-driven business process automation, focusing on organizational expertise leverage and continuous knowledge transfer across the company’s workforce. The software helps companies increase productivity by connecting people with information through cutting-edge workflow technology; attract better talent, retain employees, ensure knowledge lock within the organization to protect itself from key employees departure, and streamline information collection, analysis, and timely outcomes delivery.

Trisk’s customers range from SMBs to global enterprises, but they all depend on expertise to deliver core offerings, such as financial services. Trisk enables these teams to digitize and automate internal knowledge, so they can streamline their workflows without adding staff or increasing costs.

The Challenge

Trisk’s product roadmap included AI capabilities to enable customers to use conversational AI for workflow design, task management automation, translating information into actions, and data accessibility. Initially, Trisk relied on OpenAI to manage its AI models. However, OpenAI introduced several operational difficulties:

• Unstable Updates and Policy Changes: OpenAI frequently updated its models and policies, which often led to disruptions in service. These changes were unpredictable, and each new update required extensive retesting and reconfiguration of Trisk’s workflows to ensure compatibility. This resulted in downtime, affected client experiences, and increased the operational burden on Trisk’s development team.
• Data Privacy and Security Concerns: Because Trisk serves customers in regulation-heavy industries like finance, data security is crucial. OpenAI’s policies around data handling and storage were not always transparent or robust enough to meet Trisk’s strict compliance requirements, including SOC 2 Type 2. This uncertainty raised concerns about data protection, especially when dealing with sensitive customer information.
• Limited Customization and Scalability: Although OpenAI provided general-purpose AI models, it lacked the flexibility Trisk needed to customize and scale specific AI models to accommodate customers’ unique workflows. OpenAI’s limitations in supporting custom models and its reliance on shared infrastructure made it difficult for Trisk to offer industry-specific AI solutions at scale.
• Performance Bottlenecks: As OpenAI’s infrastructure became increasingly popular, Trisk noticed performance issues during periods of high demand. The shared infrastructure meant that peak usage by other customers affected the performance of Trisk’s own operations. The unpredictability led to slower response times, particularly during high-volume periods, impacting user experience.

These challenges made it clear that continuing to rely on OpenAI would hinder Trisk’s ability to scale, deliver reliable service, and meet its clients’ stringent compliance and security needs. Trisk needed a better AI infrastructure, prompting them to seek a new solution. The team was especially impressed by AWS and saw an opportunity to streamline their business and run their entire operation on AWS services.

“The platform is now incredibly powerful, but what’s even more important from a customer-facing standpoint is that we have a robust, stable, reliable system that will not fail.” – Vlad Anikin, Chief Technology Officer at Trisk

The Solution

Trisk chose to partner with AllCloud to design and build a solution that would provide advanced machine-learning models and cloud services on a secure, scalable, and reliable platform. The solution called for building an AI-driven architecture using AWS Bedrock, Lambda, and other AWS services for automating workflows and document processing. 

The Trisk team sought a robust and reliable toolset to deliver a stable customer-facing functionality, not only to replace the already built features, but also to scale the AI footprint in the product while mitigating associated infrastructure risks.

AllCloud worked hand-in-hand with Trisk to ensure the solution was tailored to meet their unique needs, including advanced compliance, security, and AI model management. AllCloud also provided expert guidance on fully leveraging AWS’s AI and machine learning services. The solution involved several components to achieve Trisk’s goals for the project:

AWS Bedrock for AI Infrastructure

A key aspect of the solution was the migration of Trisk’s AI workloads to AWS Bedrock. AllCloud facilitated this migration, enabling Trisk to access cutting-edge machine-learning models and operate in a stable, scalable environment. AWS Bedrock provides Trisk with the flexibility to use multiple AI models, including Anthropic models for text processing, and to manage their AI operations more efficiently. AllCloud’s expertise in AWS architecture ensured the migration was smooth and efficient, reducing the operational disruptions that Trisk had previously faced with OpenAI.

AI-Powered Workflow Automation

Trisk’s new AI assistant enables customers to use conversational AI to design and manage workflows without writing code. This no-code approach empowers businesses of all sizes to streamline their operations without technical expertise.

AI-Driven Reporting and Insights

The team implemented an AI-based reporting system that allows users to access and analyze critical business data. Trisk’s customers can use this feature to query outstanding tasks and stalled projects, with AI making suggestions for how to address bottlenecks. This system helps unlock delayed projects and ensures faster turnaround times.

Email Triage Automation

Email workflows enable Trisk’s customers to manage email-heavy processes, such as security incident responses, without manual intervention. An AI-powered email triage feature ensures that relevant information is automatically routed to the correct workflows, improving efficiency and reducing the chances of human error. This automation tool also leverages existing client workflow information by seamlessly integrating with Trisk’s data storage service to ensure the AI output is even more accurate.

Automation Tools and Services

Multiple AWS services power Trisk’s platform:

• Amazon ECS and EC2: Enable the deployment of scalable, containerized applications that run the core components of Trisk’s AI and workflow management systems.
• Lambda: Automates key workflows within the Trisk platform, allowing AI tasks to be seamlessly integrated with core business processes.
• Amazon Textract: Extracts and processes data from physical documents like contracts, enhancing the platform’s ability to handle complex workflows.
• DocumentDB, Elasticsearch, SQS, and SNS: These services form the backbone of Trisk’s document management and messaging systems, allowing for efficient task execution and data storage.

Security and Compliance

Trisk’s platform now meets the highest standards for security and compliance. As a SOC 2 Type 2 certified solution, the platform tracks every action taken by users, providing full transparency and audit trails. Trisk’s solution offers secure document management, including watermarking and tracking, to protect sensitive data.

By leveraging AllCloud’s deep expertise in AWS architecture and cloud migration, Trisk was able to build a scalable, secure, and AI-driven platform that significantly improved both internal operations and customer-facing processes.

The Results

The project has had a transformative impact on Trisk’s internal operations and strategic growth. The company was able to achieve the following outcomes:

• Stronger Strategic Positioning and Higher Win Rate: By integrating AI into its products, Trisk has strengthened its competitive differentiation. The company has enhanced its ability to deliver innovative AI-driven solutions, allowing Trisk to position itself as a leader in business process management. This has resulted in significantly increased attention from prospects and a higher win rate for the sales team.
• Streamlined Internal Processes: Trisk’s internal operations have become more efficient through the use of its AI-powered platform. Key internal functions, such as documentation management, service desk management, and reporting, have been automated, allowing Trisk’s employees to focus on higher-value tasks rather than routine manual processes.
• Improved Productivity Across Teams: AI-based reporting tools help Trisk optimize its internal workflows, identifying bottlenecks and stalled projects. This enables Trisk’s teams to accelerate project completion and reduce downtime, ultimately contributing to faster delivery of client projects and better internal alignment.
• Enhanced AI Capabilities: Integrating AWS Bedrock and other AI tools has enhanced Trisk’s internal AI capabilities. It has also laid the groundwork for future AI-driven innovations.
• Future Growth: Trisk has not only stabilized its platform through its migration to AWS but has also set the stage for future scalability. An upcoming integration of SageMaker and other planned AI initiatives position Trisk to extend its offerings, making it more competitive in new markets and expanding its reach into sectors like M&A and advanced financial analysis.

Vlad Anikin, Chief Technology Officer at Trisk, explains, “AWS services and AllCloud have provided a completely different level of managing AI processes, compliance, data security, and ethical use of the data. This is a new level of workflow automation on the market, powered by actionable information access, exceptional accountability and compliance in a highly regulated environment, ethical AI use, and a lower barrier of entry for non-tech savvy business people. But what’s even more important from a customer-facing standpoint is that we have a robust, stable, reliable system that will not fail.”

Learn more about how Trisk is providing organizations with higher-quality operations with AI-driven Conversational Process Automation. Visit www.trisk.io