In this Leading with Data session, we dive into the journey of Anand Ranganathan, a visionary in AI and machine learning. From his early days at IBM to co-founding innovative startups like Unscramble and 1/0, Anand shares insights into the challenges, transformations, and future of AI. Join us as we explore his entrepreneurial experiences, the impact of deep learning, and his vision for the future of AI and its applications.
You can listen to this episode of Leading with Data on popular platforms like Spotify, Google Podcasts, and Apple. Pick your favorite to enjoy the insightful content!
Key Insights from our Converastion with Anand Ranganathan
- Balancing symbolic AI and deep learning is key for precise reasoning in specific domains.
- Deep learning’s rise demands agility in product development and market strategies.
- AI services companies focus more on customer relationships and tailored solutions than product firms.
- Agentic workflows will transform AI integration, but human-AI collaboration boundaries need clarity.
- For AI/ML careers, domain expertise and staying updated are essential for success.
- AI’s future will reshape software engineering, requiring continuous learning and adaptation.
- Domain knowledge is vital as AI disrupts generic software engineering roles.
Let’s look into the details of our conversation with Anand Ranganathanl!
How did your journey in AI and ML begin, and what were the early days like for you?
My journey in AI began with my PhD at the University of Illinois, where I delved into the intersection of AI and distributed systems. Back then, AI was more about symbolic or logical reasoning, quite different from today’s landscape. I worked on AI planning, which involves transitioning the world from one state to another using a set of actions. After my PhD, I joined IBM Research, where I tackled big data problems and was part of the team that built IBM’s stream processing offering. It was an era dominated by classical AI, but as deep learning gained traction in the 2010s, the field transformed dramatically.
What motivated you to leave IBM and start your own venture?
After a decade at IBM, I was eager to tackle interesting problems that I identified in the industry. Meeting the right people who shared my vision and spotting a market opportunity were the catalysts for me to co-found my first startup, Unscramble. We aimed to be nimble and innovative in solving challenges, which was a different experience from the corporate environment at IBM.
Can you explain the two different problems Unscramble focused on, and how they were connected?
Unscramble initially tackled real-time streaming data problems, specifically in the telecommunications sector. We then realized there was also a need for analytics on historical data. Although the domains were different, the underlying commonality was in queries on structured data and triggers on streaming data. Our solutions ranged from natural language queries on databases to defining marketing campaigns in real-time using a natural language interface.
How did the rise of deep learning impact your products at Unscramble?
Deep learning’s rise was significant, especially for our natural language to SQL translation product. We had to evolve our techniques as deep learning models became more adept at handling such tasks. Eventually, when fine-tuned SQL generation models emerged, it was clear that the space was being disrupted. We were already exploring an exit strategy, and the timing worked out for us to sell the product before the disruption became too great.
What are the differences between running a product company like Unscramble and a services company like 1by0?
Running a product company is about showcasing what you have and adapting it to customer needs, while a services company is about understanding the customer’s problem and crafting the right solution. At 1by0, we focus more on account and project management, certifications, and maintaining close partnerships with vendors like AWS and Databricks. It’s a different trajectory, with a stronger emphasis on customer relationships and delivering tailored solutions.
Reflecting on your entrepreneurial journey, what are some key learnings and things you might do differently?
One key learning is the balance between tackling interesting problems and focusing on market demand. At Unscramble, we sometimes prioritized interesting challenges over market viability, which, while intellectually satisfying, wasn’t always optimal for startup growth. In the services space, the challenge is deciding how much to invest in exploratory solutions versus safer, well-understood ones.
How do you envision the future of AI, particularly in the context of symbolic AI and deep learning?
I believe there’s a need for a balance between symbolic AI and deep learning, especially in domains requiring precise reasoning, like medicine. While LLMs are improving in reasoning capabilities, there’s still a need for provable and accurate knowledge, which symbolic AI can provide. Breakthroughs in simplifying the construction of knowledge bases could be key to advancing symbolic AI.
What trends do you foresee in AI in the near future, and how do you see agentic workflows evolving?
Agentic workflows are gaining traction and will continue to do so. They offer a way to integrate AI into everyday work more seamlessly. However, the boundary between human and AI collaboration is still fuzzy. Deciding when AI can take action automatically and when to involve a human will be critical. I also see AI becoming more embedded in software development, changing the skill set required for software engineers.
What advice would you give to those just starting their careers in AI and ML?
Focus on gaining domain expertise in addition to technical skills. Domain knowledge is less likely to be disrupted and can complement your technical abilities. Stay abreast of advancements in AI and experiment with different tools and frameworks to enhance your effectiveness. It’s a rapidly changing field, so continuous learning is essential.
End Note
Anand Ranganathan’s journey reflects AI’s rapid evolution and potential. From IBM to pioneering startups, his story underscores the importance of adaptability, domain expertise, and balancing innovation with market needs. As AI reshapes industries, his insights highlight the critical role of human-AI collaboration and continuous learning. The future of AI is exciting, and leaders like Anand are paving the way for transformative advancements.
For more engaging sessions on AI, data science, and GenAI, stay tuned with us on Leading with Data.
Check our upcoming sessions here.
Hello, I am Nitika, a tech-savvy Content Creator and Marketer. Creativity and learning new things come naturally to me. I have expertise in creating result-driven content strategies. I am well versed in SEO Management, Keyword Operations, Web Content Writing, Communication, Content Strategy, Editing, and Writing.
