Moveworks is a conversational AI platform that connects people and the systems we use to get work done. It’s powered by some of the world’s most advanced Large Language Models — including OpenAI's GPT-3 — to resolve employees’ support issues end-to-end. By engaging in a simple conversation with a chatbot, employees can instantly address common IT issues, HR requests, and more, without the intervention of a help desk agent.
Challenge: Streamlining ticket resolution for enterprise complexity
The Moveworks Enterprise Answers skill allows users to search any snippet of information buried across different applications and systems — all from one place. To enable this, they ingest and analyze thousands of resources per customer across various integrations and domains — this includes knowledge base articles, company policies, user permissions, service catalog forms, and FAQs. Customer Success and Technical Support reps must also understand this growing web of constantly changing resources to effectively troubleshoot customer issues, while engineering teams optimize and enhance the intelligence engine. In line with keeping a comprehensive data privacy approach, these resources are abstracted through internal APIs and stored in AWS, making it difficult for frontline teams to make sense of the complexity in a compliant manner. Some challenges that stem from this include:
- Inability to search across customer resources to find troubleshooting information (e.g., last time a resource was updated), which is critical for understanding a customer’s ingestion pipeline health.
- In order to debug issues with resources, tools had to exist to deeply inspect them:
• For articles, this meant accessing the underlying HTML to understand snippet extraction.
• For service catalog items, this meant viewing what fields/components were available on a form to ensure a high quality user experience when filling forms in chat.
- Comparing resource history across versions was cumbersome - preventing the team from unblocking a customer issue over what was often a simple configuration change on the customer side.
As a result of these challenges, finding the root cause of customer questions like “Why can’t this new hire access this article in the bot?” or “Why did the UI policy on this catalog item change?” became a laborious process and support ticket resolution time suffered.
Solution: Consolidate customer resource data into a single Superblocks UI
With Superblocks, the Moveworks’ Customer Success Engineering team was able to build an application in a few days to enable its Customer Success and Support teams to troubleshoot support issues faster and increase their customer CSAT, without the help of their Search and ML Engineering teams.
The Resource Viewer Application provides Customer Success and Support agents with an interface to quickly understand all resources for a customer, filtered by type (knowledge article, form, FAQ) as well as any custom search parameters. With easy to use UI filters in place, data is fetched from Moveworks’ backend S3 indices, and returned to the Superblocks frontend to display a summary of important troubleshooting information:
- Resource status and integration source (ServiceNow, Jira, Confluence, etc)
- Latest ingested / indexed timestamps
- Ingestion / indexation failures
Immediately, the team knows if a resource is being served properly and whether or not it’s up to date.
Should they need to dig in further to see the contents of a resource - perhaps a geo identifier is present in the article, which affects the personalization algorithm and might change how content is served to end users in different regions - that level of detail is just a click away. No need for a CLI or manually accessing files from S3.
Lastly, the Diff Viewer component displays a side by side view of resource versions from two ingestion dates so Customer Success and Support can compare any historical resource snapshots, just like code diffs in GitHub. In seconds, they can identify the root cause for unexpected changes on an ingested resource from the customer’s knowledge base, an analysis that previously could have taken 30-plus minutes of the Search Engineering team’s time.
1. Support of common Python libraries and custom packages
The Customer Success Engineering team was already familiar with writing Python and utilizing internal gRPC services. Superblocks’ support for Python and custom packages allowed them to easily port these over to Superblocks where they could build a dedicated UI on top of their business logic.
2. Ability to build complex & secure UIs without React expertise
With drag and drop frontend components, Superblocks turned engineers without full stack expertise into engineers who could create entire user interfaces connected to their backend AWS services, in just a few hours. Notably, the pre-built Diff Viewer component provided a key visualization for customer facing teams to troubleshoot resource changes fast, and saved the dev team many days from building something in-house.
3. Secure and least-privileged access to customer data
The On-Premise Agent allowed Moveworks to build tools in Superblocks that connected to its internal services securely from an agent deployed within its AWS VPC. This was the only way for Superblocks to securely interact with customer data while abiding by Moveworks security standards, while RBAC support ensured data was accessed on a least-privileged basis. At the same time, Moveworks received the latest features and security updates from the Superblocks Cloud platform.
Results and future state
With Superblocks, Moveworks has created tools that allow its Customer Success and Support teams to move faster and more efficiently, reducing support ticket resolution time and increasing customer CSAT significantly. Moveworks plans to build more Superblocks tooling for its teams to gain more visibility around its platform and customer configurations. As their enterprise user base continues to grow, these tools for support and onboarding will be crucial to solve issues efficiently, remove friction, and deliver delightful user experiences in the workplace.
Get the latest Superblocks news and internal tooling market insights.