Last Updated on August 13, 2024
Join us as we discuss how Prakash has harnessed the power of AI to improve productivity in the biotech industry, reduce default rates in the financial sector, and shape the future of the business landscape. Drawing from his diverse background, which spans technology, finance, infrastructure, and energy, Prakash offers invaluable insights into the potential of AI to revolutionize various industries and redefine the way we address complex challenges.
In our conversation, we will touch on his work at Sale Tech Singapore, where he pioneered the first real-time machine learning credit scoring system in Indonesia, and FamiliarAI, a cutting-edge natural language processing application designed to enhance the productivity of PhDs in the biotech sector. We will also explore his thoughts on the broader implications of AI for society, the ethical considerations surrounding its use, and his vision for the future of AI-driven solutions.
Don’t miss this exclusive opportunity to learn from Prakash Narayanan, a true trailblazer in the field of AI, whose multifaceted expertise and experience make him uniquely positioned to offer thought-provoking insights into the transformative potential of artificial intelligence. Prepare to be inspired, informed, and challenged as we navigate the frontiers of AI innovation together.
What inspired you to transition from finance to technology and AI, and how has your background in finance informed your work in technology?
I started my career in tech, moved to finance, and back to tech. My experience in finance gave me a strong foundation in strategy, financial analysis, and risk management, which are all critical skills for any entrepreneur. Finance provided me with a general understanding of the living, breathing entity that is the corporation.
However, what ultimately inspired me to transition from finance to technology and AI was the desire to work on cutting-edge solutions that have the potential to make a significant impact on people’s lives. I saw an opportunity to leverage my skills and expertise to create innovative solutions that could solve some of the most challenging problems faced by businesses and society at large.
In particular, my interest in AI was sparked by the realization that it has the potential to transform virtually every industry by providing advanced tools for data analysis, prediction, and automation. By building FamiliarAI, I am able to apply my skills and experience to create AI-based tools that can improve the efficiency and productivity of technical sales teams in biotech and other industries.
How did you identify the need for a natural language processing application to improve technical sales, and what were some of the challenges you faced in building FamiliarAI?
The idea for FamiliarAI came about serendipitously when I was speaking to a biotech CEO about their struggles at the time and the effort required for their technical sales team to read and analyze research papers. They recognized that this was a significant bottleneck in their sales process and asked if there was a way to improve the efficiency and productivity of their team.
While building Sale Tech, we’d built a lot of tools to read loan applications and extract data to be used for credit scoring, and I was surprised to learn that such tools were not in use in biotech yet due to domain expertise required to understand the material. Reading credit reports in Indonesian language turns out to be not dissimilar to reading cancer research – both require machine learning to form a structure and context for the field before performing inference.
This led me to develop FamiliarAI as a software-as-a-service (SaaS) platform that uses natural language processing (NLP) to analyze, summarize, and create contextual AI-generated text for technical sales.
Building FamiliarAI presented several challenges, including identifying the right data sources, choosing accurate and reliable NLP algorithms, and ensuring that the platform could be easily integrated into existing workflows. To overcome these challenges, we worked closely with our clients to understand their specific needs and develop customized solutions that met their requirements. We also invested heavily in building expertise in NLP, machine learning, and software engineering to ensure that we had the technical expertise required to build a best-in-class platform.
Can you walk us through your experience building Sale Tech and creating the first real-time machine learning credit scoring system in Indonesia? What impact do you think this technology will have on the financial sector in Indonesia and beyond?
Sure! When I founded Sale Tech in 2018, I saw a need for more efficient and accurate credit scoring in Indonesia’s financial sector. Traditional credit scoring methods were often slow and relied on limited manual data sources, making the credit analysis process relatively expensive. Lenders have to recover this expense by charging borrowers higher interest rates, making credit unnecessarily expensive.
Mobile phones generate a lot of data, none of which was being used to measure creditworthiness in Indonesia. I believed that machine learning algorithms could be used to process large amounts of data in real-time and generate more accurate credit scores, which could decrease the cost of issuing loans and thereby expand access to credit for more people.
The initial challenge was to gather and process the right data sources, which included both traditional financial data and non-traditional data from sources such as social media and mobile phone usage. We also had to make sure that our system complied with local regulations and data privacy laws. We then used machine learning algorithms to analyze this data and generate credit scores in real-time.
The impact of this technology has been significant. By providing more accurate and real-time credit scores, we were able to reduce default rates and expand access to credit for more people. This has helped to fuel economic growth in Indonesia and beyond, as more people are able to access the capital they need to start businesses and invest in their futures.
As a multi-disciplinary engineer, how do you see AI impacting various industries in the future, and what opportunities do you see for entrepreneurs to innovate with this technology?
AI plus human oversight is going to be amazing in many industries. We saw it in credit in Indonesia, where a combination of machine learning with deep local credit expertise delivered far superior results to either alone.
I used to remember phone numbers growing up, and now every phone number is on my phonebook on my phone. Modern AI is going to acquire mini skills that way, from writing emails to planning corporate strategy.
If you’re an entrepreneur, first learn what AI capabilities are. Play around with Stable Diffusion and ChatGPT. Learn how to use them. Then look around for places where applying the technology can allow you to 10x quality, speed, and productivity. Literally, every single segment of our lives will be permeated by AI within 5 years. But it still takes skill to identify the most natural vectors of application.
Can you discuss your approach to risk-taking and how you manage uncertainty in business? How do you balance innovation with practicality and ensure that your ventures are financially viable?
My approach to risk-taking is based on hypothesis testing. I try to avoid getting caught in the typical investment banker/consultant cost-benefit analysis approach, as at the early stages, this leads to paralysis due to a lack of information on new markets and products. Instead, I use the lean startup methodology, which is to identify the unknowns at each state in the business, and build small low-risk tests of the hypothesis in order to clarify.
This leads to greater speed of decision-making and iteration. At the end of the day, in a startup, you have infinite upside and limited downside. So, you have to go ahead and take the plunge, fail a little, and iterate.
Financial viability is really a balance between growth, profits, and fundraising. Knowing when to do what and timing it are super important.
How do you build diverse teams and approach cross-cultural communication in your work, given your experience working in different cultural environments and speaking multiple languages?
I have learned to listen, trust and nudge. If you hire well, most people are motivated to do great work. If you put your faith in them, give them your trust, and be clear in your expectations, they can perform extremely well. After some time, a good team is a well-oiled machine that works even when you don’t pay attention because everyone knows exactly how you would react in a situation.
Cross-cultural teams can be awesome because everyone is keen to learn from a different culture if they have the space to grow. I always plan a tolerance for losses and delays, make plans way, way ahead of time, and draw up decision trees on what to do if something happens. If you leave enough leeway for these things, place expectations on people to get things done regardless of language or cultural barriers, and keep out of their way, they eventually figure it out. It’s important not to intervene to just get tasks finished quickly, as there needs to be a burn-in time for acclimatization.
Can you share some of your strategies for scaling businesses, and how you determine which industries and opportunities are worth pursuing?
I have a very video-gamey view of the world, where there are certain things you do as you level up. There are certain power-ups you need to get like important hires, and leveling up means you get to tackle bigger challenges by serving bigger clients. So, in terms of scaling, you could say I try to clear each level first, then go on to the next level.
Target opportunities emerge from having the right skills, the right market at the right time. Sometimes, you spend time gaining skills but wait to deploy because of timing. At the end of the day, you have to tackle a sufficiently big challenge that you’re willing to commit a decade of your career to see it through. These are not easy to identify, so doing small projects to test markets is a very important part of the discovery process.
What advice do you have for aspiring entrepreneurs who are looking to build successful businesses, particularly in industries that are rapidly evolving due to technological advances?
My advice to aspiring entrepreneurs is to start by identifying the most significant pain point that exists for a single customer in the market. And then work to figure out an excellent solution that absolutely nails that problem, and then work on generalizing that solution.
Always be aware that technology is useless if it does not solve problems. A hammer without a project is just a waste of space. There is no competition. There is only a customer with a pain point. If the customer had a solution, they would not be in pain.
This viewpoint takes away the whole “I can’t compete with Google” issue. There are always people with problems because large firms need to build tools that generalize and small firms can always afford to build specific tools that work extremely well for a small subset of customers.
Finally, just start. As Mike Tyson said, “Everyone has a plan until they get punched in the mouth.” You only find out the issues when you start, and then the game’s afoot as you pivot to find product market fit and scale.
