Last week Seek joined Snowflake for an episode of their AI Builders series, with our Founder & CEO Sarah Nagy to discuss how Seek AI's natural language interface revolutionizes data accessibility, empowering users to extract insights in real-time from Snowflake, fostering trust through advanced models and collaboration.
In the episode we covered:
- Seek AI solutions
- A live demo
- Hallucination protection
- Seek AI's future vision
Catch the full episode below for an immersive journey into the intersection of AI and Snowflake, shaping the future of data-driven insights.
Don’t have time to watch? Here’s a summary of the video for reference.
The Origin Story of Seek
Former data scientist Sarah Nagy started Seek to increase the effectiveness of data and solve the widespread problem of data bottlenecks within businesses. As a data scientist Sarah wanted to focus on deep research on data to find valuable insights that could add value for the business in the form of profitable investments. Instead, non technical users would need her help answering ad hoc data questions because the tools they had weren’t serving their needs. This resulted in Sarah having to write a lot of manual, repetitive code every day for her colleagues. In 2020 Sarah came across LLM’s (Large Language Models) that could write code efficiently and quickly. She realized that non technical users could ask these models questions, and the model could replicate the same code that Sarah had to frequently write for her colleagues. This realization led to the founding of Seek.
What is Seek?
Seek is a natural language interface for structured data that helps business users ask ad hoc data questions and get answers back in real time.
What pain point is Seek solving?
Seek helps remove the data bottleneck that so many businesses suffer from today, with limited and strained data analyst resources unable to answer ad hoc questions inbound from all sides of their company. Time is returned to both the data team and business users resulting in greater bandwidth for business impact. Ultimately, the insights Seek helps to extract makes company data more effective.
What kind of questions can Seek answer?
Seek can transform plain language questions about data into an automated query of the data that returns real insights. In this demonstration, Seek queries a data warehouse to get answers about week over week sales growth of particular brands.
How accurate is Seek?
Seek is powered by the most accurate natural language to data model, having performed with over 90% accuracy on Yale’s Spider Leaderboard test benchmark.
What is Seek planning for the future?
Seek is proud to be a Snowflake Snowpark Container Services partner and is excited to deliver unprecedented value to enterprise customers using Snowflake moving forward. Seek plans to expand from natural language to SQL querying into other coding languages like Python.
Interested in learning more about Seek’s data accessibility revolution? Book a demo today: https://www.seek.ai/trial
