The Center for Data Innovation spoke with Chun Jiang, co-founder and CEO of Monterey AI, a customer feedback analytics platform based in San Francisco. Jiang explained how businesses can leverage natural language processing to better gain insights from customer feedback.
Morgan Stevens: What motivated you to start Monterey AI?
Chun Jiang: I was motivated to start Monterey AI because it addresses two key problems that I encountered in my previous roles at Uber, Scale AI, and Foursquare. Firstly, manually reviewing and categorizing large quantities of user feedback and support tickets is inefficient and impractical. Secondly, valuable user insights are scattered across various platforms such as social media, CRM systems, and communication tools like Slack and Discord. It is difficult to keep track of everything and restrict data access to just one person or team within a company. Monterey AI aims to solve these challenges.
Stevens: Can you explain the technology behind Monterey AI’s copilot?
Jiang: Monterey AI’s technology utilizes a sophisticated system that combines a wide range of natural language processing (NLP) techniques. This includes both traditional methods and cutting-edge language models. By employing this versatile approach, Copilot can effectively handle different tasks related to customer data analysis. Task-specific models enable Copilot to comprehend and process incoming data, performing tasks such as sentiment analysis, content retrieval, and classification. This allows businesses to gain valuable insights from customer feedback and communication. Monterey AI also incorporates embeddings and advanced search capabilities, providing robust query and chat interfaces to enhance the user experience.
Additionally, the system organizes incoming data thematically, making it easier for organizations to identify recurring themes and topics within customer feedback. Overall, Monterey AI’s Copilot technology offers a comprehensive solution for businesses seeking to leverage NLP and advanced AI techniques to better understand and respond to customer voices, ultimately improving their products and services.
Stevens: How does Monterey AI customize its copilot to cater to the specific needs and nuances of different industries?
Jiang: Monterey AI offers proprietary solutions that can be quickly deployed and adapted to specific industry contexts. To customize its copilot, Monterey AI develops industry-specific models, employs advanced natural language understanding (NLU) techniques, and continuously learns from data. This enables the delivery of tailored product insights for clients in various industries.
Stevens: What are the biggest challenges to automating qualitative data analysis?
Jiang: Automating qualitative data analysis faces significant challenges due to the unstructured nature of the data and the nuances of human language. These challenges include dealing with the lack of predefined structure, accurately capturing contextual information and nuances, addressing subjectivity and ambiguity in interpretation, ensuring scalability for large datasets, and obtaining high-quality annotated training data. Furthermore, generalizing automated approaches across domains and languages, validating and interpreting results, addressing ethical concerns related to biases, and continuously adapting to evolving language trends and contexts are persistent challenges in this field. Despite these obstacles, advancements in natural language processing (NLP) and machine learning offer opportunities to improve automated qualitative data analysis while recognizing the ongoing importance of human expertise in validation and interpretation.
Stevens: What does the future look like for Monterey AI?
Jiang: Monterey AI is at the forefront of delivering autonomous product solutions. Our initial focus is on providing the best tool for analyzing qualitative data. However, we are actively working towards a future where we offer the best AI analysis engine for both qualitative and quantitative data, and power other AI solutions in areas such as automated bug fixing and UI design suggestions.