From NLP PhD to Y Combinator Startup: My Journey to Revolutionizing Customer Feedback Analysis

From NLP PhD to Y Combinator Startup: My Journey to Revolutionizing Customer Feedback Analysis

Table of Contents

1. Introduction
2. My Journey to Becoming a CEO of a Funded Startup
3. Making a Transition from Linguistics to AI
4. The Inspiration Behind my Research on Natural Language Processing
5. The Significance of Knowledge Representation in NLP
6. The Role of Commercialization in Turning Knowledge into Profit
7. The Creation of Somatic: Analyzing Customer Feedback
8. Challenges of Analyzing Customer Feedback Using NLP APIs
9. The Unique Approach of Somatic in Analyzing Feedback Data
10. The Power of Human-in-the-Loop Editing in Theme Analysis
11. The Importance of Visualizing Feedback Data
12. The Competitive Advantage of Somatic
13. Lessons Learned and Advice for Starting a Company
14. Conclusion

🚀 My Journey to Becoming a CEO of a Funded Startup

In this article, I'm excited to share my story about how I ended up becoming the CEO of a funded startup called Somatic. I'll take you through my journey from studying linguistics to entering the world of artificial intelligence and natural language processing (NLP). I'll also discuss the commercialization of NLP research and how it led me to create Somatic, a software-as-a-service company that specializes in analyzing customer feedback.

Introduction

Let me start by giving you a bit of background. My passion for NLP began when I attended a conference where I learned about the field of computational linguistics. This discovery completely changed my career trajectory as I realized the potential of using computers to understand language. I was initially studying linguistics in Ukraine, but this new field sparked my interest, pushing me to pursue a career in NLP.

My Journey to Becoming a CEO of a Funded Startup

After completing my studies and obtaining my master's degree in Germany, I made my way to New Zealand. It was there that I pursued my PhD at Waikato University, focusing on keyword extraction and knowledge representation in text analysis. This research became the foundation for my work at Somatic, where we analyze customer feedback to identify the most common themes.

Making a Transition from Linguistics to AI

My journey from linguistics to AI was not a direct path. I started my career in academia, but I soon realized that I wanted to make a more significant impact by bringing NLP research into the business world. I joined a company called Pingar, which was one of the few companies working on NLP in New Zealand at the time. However, my time there taught me valuable lessons about sales and the importance of understanding customers' needs.

The Inspiration Behind my Research on Natural Language Processing

During my academic years, I was continuously inspired by the power of NLP and its ability to transform the way we analyze and understand text. My research focused on extracting keywords and identifying common themes in text analysis. I was fortunate to have the opportunity to use open-source projects like Maui and Wikipedia to enhance my research. These projects allowed me to create controlled vocabularies and knowledge representations to improve the accuracy of theme extraction.

The Significance of Knowledge Representation in NLP

Through my research, I discovered the importance of knowledge representation in NLP. Traditionally, controlled vocabularies, like Agrovoc, were used to map different terms to a universal theme. However, I found that Wikipedia offered a more comprehensive and crowd-sourced alternative. By leveraging Wikipedia's vast collection of articles, I was able to improve the accuracy of theme extraction by disambiguating WORD meanings and selecting Relevant articles.

The Role of Commercialization in Turning Knowledge into Profit

Upon completing my PhD, I realized that academia was not for me. I craved the opportunity to bring my research into the business world and make a tangible impact. I joined a company called Pinger to explore the commercialization of NLP technology. We initially focused on automating report generation and language analysis using NLP APIs. However, we quickly pivoted to offering metadata solutions for document management systems like SharePoint.

The Creation of Somatic: Analyzing Customer Feedback

After leaving Pinger and taking time off for maternity leave, I knew I wanted to start my own company. With my expertise in NLP and natural language analysis, I identified the need for a solution that could analyze and extract Meaningful insights from customer feedback. This led me to create Somatic, a software-as-a-service (SaaS) company that specializes in analyzing customer feedback to identify the most common themes and extract valuable insights.

Challenges of Analyzing Customer Feedback Using NLP APIs

When developing Somatic, I encountered several challenges in analyzing customer feedback using existing NLP APIs. Many of these APIs were designed for analyzing news articles, not the unique language used in customer feedback. The results were often inaccurate and lacked the relevant themes needed for effective analysis. This is why I decided to develop Somatic from scratch, specifically tailored for customer feedback analysis.

The Unique Approach of Somatic in Analyzing Feedback Data

The competitive advantage of Somatic lies in its ability to analyze customer feedback without the need for training specific models. Our system can analyze any set of feedback and extract relevant themes, providing fast and accurate results. The themes are then further refined by an editor who can merge, delete, or modify them for better accuracy. This human-in-the-loop editing process ensures the highest level of specificity and relevance in the analysis.

The Power of Human-in-the-Loop Editing in Theme Analysis

One of the key features of Somatic is the ability for users to edit and refine the analysis results themselves. Using a simple drag-and-drop interface, users can organize and modify the themes according to their specific needs. This allows for greater control and customization, ensuring that the analysis accurately reflects the insights they are seeking.

The Importance of Visualizing Feedback Data

Analyzing customer feedback is essential, but visualizing the data is equally important. Somatic provides a range of visualization options, allowing users to filter and explore the feedback data based on specific themes or sentiments. These visualizations provide valuable insights into customer trends, sentiment changes, and the impact of various initiatives on customer experience.

The Competitive Advantage of Somatic

The competitive advantage of Somatic lies in its unique approach to analyzing customer feedback. Unlike other NLP APIs and solutions, Somatic does not require training models or predefined taxonomies. Instead, it analyzes the data dynamically, allowing for greater flexibility and accuracy. This enables businesses to gain valuable insights from their customer feedback quickly and effectively.

Lessons Learned and Advice for Starting a Company

Throughout my journey as a CEO and entrepreneur, I have learned valuable lessons that I would like to share. Firstly, it's essential to have a clear understanding of the problem you are solving and who your target audience is. Secondly, being open to feedback and continuously adapting your product or service based on customer needs is crucial for success. Lastly, building a strong network, seeking mentorship, and staying resilient are key qualities every entrepreneur should possess.

Conclusion

In conclusion, my journey from linguistics to becoming the CEO of Somatic has been an exciting and challenging one. The field of NLP has immense potential, and I am grateful to have the opportunity to contribute to its growth. With Somatic, we aim to revolutionize customer feedback analysis and provide businesses with valuable insights to enhance their customer experience.

🚀