Explore the future of ai with Dr. Vikas Agrawal, Senior Principal Data Scientist at Oracle Analytics Cloud. In this Leading with data session, shares insights on problem solving in data science, MLops, and the impact of generative ai on business solutions. The discussion ranges from practical approaches to pitfalls in data science projects and offers essential advice for aspiring data scientists.
Key insights from our conversation with Vikas Agrawal
- In data science, focusing on understanding the problem is crucial and takes the majority of the effort.
- A successful proof of concept (POC) in data science must consider not only the technical aspects but also the practicality and scalability of the solution.
- Clear communication and setting realistic expectations with customers are vital to avoiding costly misunderstandings driven by ai hype.
- Generative ai has the potential to revolutionize business solutions, especially in areas related to text and user interfaces.
- Developing a career in data science requires a strong foundation in mathematics and a deep understanding of algorithms.
- In enterprise environments, ensuring the trustworthiness and trustworthiness of ai results requires new validation techniques.
- As ai tools evolve, data scientists need skills to power and improve these tools, not just operate them.
Watch more interesting sessions with leading ai and data science experts on our community channel!
How do you balance technical depth with a macro view in data science?
In my daily work, I owe a lot to my mentors from various esteemed institutions and companies who instilled in me the philosophy that technology is a means to an end, not the end itself. The key is to spend a significant amount of time understanding the problem – about 90% of the effort goes into this. The rest is about finding solutions, which often involves looking at how others have tackled similar problems and what the customer ultimately needs. This approach has been instrumental in connecting technology to business impact.
What is your approach to solving a customer’s problem?
Once we’ve identified a problem worth solving, we first make sure we have the data necessary to address it. We then evaluate whether the technology exists to resolve the problem within a reasonable time frame. If we see a path, even if it is a couple of years away, we will proceed with a proof of concept (POC). This proof of concept is comprehensive, covering everything from data pipelines to end-to-end functionality, although scalability at this stage is not the main concern. The goal is to have a clear path to the algorithms, data sources, and the nature of the outcome we seek.
How do you handle the optimization phase and ML operations?
After a successful proof of concept, we enter the optimization phase, which is where most of the work lies. This involves ensuring that the model adapts to different business processes and geographies, and can be corrected only when it goes out of distribution. It’s also about ensuring that the model can be retrained efficiently and scale appropriately. This phase is critical because it is where the model goes from a concept to a practical and deployable solution.
What are the most common mistakes in data science projects?
The most costly mistakes are often related to ai overkill and miscommunication. It is essential to establish clear and mutual expectations with the client. Customers often have high expectations due to industry rumors around ai, without realizing that the latest advances may not always provide the right answers they are looking for. Another mistake is defining the problem incorrectly, either by not directly addressing the customer’s problem or by attempting to “boil the ocean.”
<h2 class="wp-block-heading" id="h-how-do-you-interact-with-generative-ai-in-your-workflows”>How do you interact with generative ai in your workflows?
Generative ai is not widely used in most companies due to concerns about copyright and intellectual property contamination. However, we take advantage of commercially available open source material. Generative ai has significantly advanced in areas such as text summarization, text augmentation, and providing explanations. Reliability remains a challenge and we are exploring techniques to filter the results of large language models (LLMs) to ensure they are reliable for enterprise use.
<h2 class="wp-block-heading" id="h-what-impact-do-you-foresee-generative-ai-having-on-enterprise-solutions”>What impact do you foresee generative ai having on business solutions?
Generative ai will likely have the most significant impact on workflows that involve text execution, such as information retrieval and user interfaces. For example, it can dramatically improve enterprise search by retrieving semantically similar snippets of text. It can also revolutionize natural language interfaces for databases, allowing users to ask questions in natural language and receive precise SQL responses.
What advice would you give to those entering the data science field today?
It’s an exciting time to be getting into data science, but it’s essential to have a solid foundation in mathematics and understand the algorithms you’re working with. As ai tools become more sophisticated, the ability to augment and improve them will be a valuable skill. Those who can create new algorithms or understand the complexities of existing ones will be in high demand.
Summary of the conversation with Vikas Agrawal
In this insightful session, Dr. Vikas Agrawal shared key insights for success in a data science career. From emphasizing understanding problems to navigating obstacles and embracing generative ai, the interview provides a roadmap. Aspiring data scientists are encouraged to build a solid foundation in mathematics and algorithms for an ever-evolving field. This interview heralds a new era of innovation in ai.
Stay tuned with us on Lead with data to catch up on the journeys of more pioneering ai and data science leaders in the industry. You can check our next Leading with data sessions here!
