Rishabh Srivastava, the co-founder of Defog.ai, is on a mission to democratise data access and analytics.
With a strong foundation in coding and machine learning, he is dedicated to using technology to make data more accessible and valuable for people. This interview will delve into Rishabh’s background and journey to building Defog.ai.
We’ll also explore the company’s vision for enabling data superpowers for everyone and the innovative tools they are developing to make working with data more seamless and efficient.
YC- Rishabh, Welcome to the Yellow Chapter. It’s great to have you here. Our goal today is to understand better who you are as a person, your life experiences and choices, and learn more about Defog.ai. Please tell us a bit about your upbringing, family, and education.
Rishabh- I was born in Raebareli, a tiny town in Uttar Pradesh, India. I was born during a period of economic liberalisation when the Indian government was creating special economic zones in the neighbourhoods of Delhi. My family and I benefited from this push for market expansion.
When I was 3, we moved to Ghaziabad, a city near Delhi, where my father worked as a factory supervisor for Samtel Electronic Devices. The company produced parts for CRT televisions. 📺📺
My mother, who worked as a primary school teacher, encouraged my curiosity from a young age. She always took the time to answer my questions, even if they seemed silly, and would research any topics she wasn’t familiar with. As a result, I developed a love for learning and exploration.
In ten years, my mom became the school principal, and my dad was a senior manager at the factory. It was incredible to witness their journey as a kid and see the transformation they could make in our lives through hard work and dedication.
Their hard work paid off, and we saw a lot of economic progress quickly. From a small one-room apartment, we could own a house and live comfortably.
My parents placed a strong emphasis on education and encouraged me to study as hard as I could. They treated my academic achievements like a social experiment and rewarded me with treats like ice cream 🍦🍦 for getting full marks on exams. However, if I scored a 98 or 99, they would ask where I lost those few marks.
My father had excellent mentorship at the factory where he worked. One of his supervisors encouraged him to buy me comic books when I was four years old to foster a love of reading 📖. When I was in the third grade, they gave me maths books for the fifth and sixth grades to challenge and push me to excel.
I did well in my studies and got through NUS.
2010-2014: Industrial and Systems Engineering, @National University, Singapore
YC- What were your engineering days like? 🎓🎓🎓
Rishabh- I joined Industrial and Systems Engineering with a business degree and was eager to excel in all my courses. Typically, students at NUS take five modules per semester, but I took on seven to learn as much as possible.
However, my grades began to decline as my interests shifted towards programming. I also dropped the degree in business after the first year as I had already taken many foundation courses.
During college, I discovered programming and the joy of building things. It was more fulfilling than striving for high grades and a high-paying job.
Coursera had just started offering free online courses, so in my third year, I stopped attending classes to focus on these courses. Even though my grades suffered, I became skilled in building software and machine learning. Looking back, I realise that I made the right choice.
YC- A lot of engineering students read my blog. Can you share five suggestions on how they can make the most of their college experience?
Rishabh-
- Focus on building things and putting them into the world rather than just learning concepts.
- Try out as many things as possible rather than optimising too early for a specific field. You might be a mechanical engineer, but you might fall in love with computing.
- Make friends who inspire you to work harder and create an impact rather than just seeking high-paying jobs.
- Get real-world experience by ensuring that the things you build are not just for personal use but are shared with the world.
- Share your work and ideas on social media and write about the cool things you are building. This will give you an edge in your career in the long run.
YC – Could you provide a brief overview of your professional journey so far?
Rishabh- Sure. I decided to do a startup post-engineering though it failed miserably.
2014: Co-founder @Quiztra – Adaptive maths learning platform.
During my final semester of college, I founded a company called Quiztra. The idea for Quiztra came about due to the difficulties I had encountered while studying mathematics.
Quiztra focused on adaptive maths learning to improve users’ mathematical skills. The platform would start with easy questions and gradually increase in difficulty as the user improved. It also provided analytics to track progress.
While the platform was popular with around 1000 users, I needed help to monetise it. I continued to work on it after college until I ran out of funds.
ICP: Initially, students at the National University of Singapore.
Learnings from Quiztra 📚📚
- Ensure that you have a strong monetisation plan in place. It’s essential to get people to use your solution and find ways to get them to pay for it.
- Consider building your business with a partner or hiring a skilled team. As a solo founder, handling everything from marketing and sales to building the platform and design can be overwhelming.
- Be aware of when it’s time to stop. I never raised any funding for Quiztra and relied on savings from my freelance jobs. By the end, I had just 29 cents left in my bank account. Don’t put yourself in that position.
YC – How did you start building tools as a data science consultant during your freelancing years from 2015 to 2017, and what was your mindset at the time?
Rishabh- I gained initial freelance customers by creating data visualisations based on publicly available information and posting them on Reddit. Millions of people viewed these visualisations, and I often received messages on WhatsApp asking about them.
This approach helped me gain exposure and establish myself as a competent and helpful resource in my industry. I started getting a lot of inbound inquiries from potential customers after including my name and email at the bottom of the visualisations.
I received requests from notable media outlets such as India Today and Times of India asking if I could create visualisations for them. Sharing my work on Reddit also helped to increase visibility and led to more opportunities.
Coding skills enabled me to build tools for companies that reached out to me efficiently. These tools were easy for companies to use. It was profitable because I would be paid — one-time costs for building the tools and recurring revenue for a low price.💰💰
Learnings as a Freelancer 📑📑
1. Networking is a vital component of freelancing success. By building relationships with other professionals and freelancers in your industry, you can improve your visibility and increase your chances of obtaining new job opportunities through recommendations from others. Don’t be afraid to reach out and make connections – it can pay off in the long run
2. As a freelancer, it’s essential to cultivate a reputation as a kind, helpful, and professional individual. Being rude or disinterested can damage your reputation and hinder your career growth. Instead, aim to be helpful and friendly to everyone you work with, regardless of their status or position. Additionally, it’s essential to be ethical, practical, and competent in your work to build a positive reputation that will pay off in the long run.
YC – Could you provide an overview of the events that led to the establishment of Loki.ai in 2017 and Popper.ai in 2020.
Rishabh- As a freelancer, I realised I frequently got the exact requests. To streamline my workflow and save time, I began exploring the development of tools and automation.
Loki.ai was founded to build data APIs by cleaning and formatting poorly structured data found on the internet.
One of our early successes was the scraping and organisation of pollution data for India. This data was initially published in disorganised HTML pages by the government. Still, we were able to transform it into clear and easy-to-use data feeds, which we then sold to businesses in India, Singapore, and Dubai.
In addition to this project, we also created machine learning tools that could generate prediction models and visualisations using a variety of data types, including political data, election results, economic data, commodity prices, and pollution data. Our focus was helping businesses make sense of complex data sets and extract valuable insights.
Popper.ai was founded to respond to the increasing demand for automated article generation, and real-time data feeds from media companies. The service provides a way for people to generate reports without spending time analysing charts, as the software handles that task. We are proud to have provided this service to several large media houses in India and have successfully served around 120M unique monthly IPs.
YC- SaaS founders have three problems — product-market fit, marketing, and sales. You have already covered the technical aspect. What’s your take when it comes to marketing and sales?
Rishabh- Creating and sharing valuable content is a critical aspect of marketing for a SaaS business.
We at Loki and Popper have found that writing about our experiences and sharing them publicly on platforms like Twitter has helped us attract potential customers. We have a blog and use Twitter to amplify the content; others have shared and republished our blog posts with permission. This allowed us to reach a larger audience without needing to hire a sales team.
Outbound marketing is an often-underrated strategy that can yield great results. This can include creating engaging content to attract potential customers, actively reaching out to potential clients through cold-calling and email outreach, and positioning oneself as a resourceful individual or expert in a specific topic.
Another crucial sales aspect is having a long-term perspective when trying to win large contracts. It means reaching out to people who may not be your target customer directly but are influential to your target customer.
For instance, when selling to government clients, we found it helpful to engage with think tanks and assist them in processing their data using our platform. This increased our brand awareness and recognition through their attribution.
YC- Can you provide more details about the discontinuation of Popper.ai and the circumstances that led to the establishment of Defog.ai?
Rishabh- One of the biggest challenges we faced at Loki and Popper was the limitations on expanding beyond our initial offerings. We had good visualisations, APIs, and dashboards, but it was difficult to address customer questions and needs outside of that scope.
However, recent advancements in AI and NLP have made it possible to interact with databases through natural language queries. This realisation led us to shift our focus from dashboards to conversational intelligence, as we saw this as a way to add significant value for our customers.
Our existing customers were excited about being able to type out questions and receive automatically generated visualised charts as answers. Through discussions with them, we confirmed that this was a valuable proposition. Additionally, this type of tool can be helpful not only for external data (data about the world) but also for internal data (data within a company), such as questions like “How are my users changing?” and “What are my highest converting channels?
We spent a year integrating our previous experience and knowledge into this new product and are now experiencing rapid growth. We were accepted into the YCombinator programme and we believe that our product has the potential to democratise data access.
YC- Could we delve into the market size, ideal customer profile, and competitive landscape in both the Indian and US markets for your product?
Market: The market for data analytics is valued at $325 billion, with self-serve analytics making up $25 billion. This market is growing at a rate of 25% per year. 📈📈
ICP: Our target market includes growth teams within large companies. These teams typically run campaigns and experiments, acquire and retain users, and analyse what’s working. The companies we target are typically large enterprises or startups at the seed to Series B stage, as they have the resources and need for our solution.
Pricing: At the moment, our pricing structure is still being refined. Our entry-level tier starts at $100/month, while our highest tier can reach several thousands of dollars per month.
Competition: Established players like Tableau and Power BI dominate the self-serve analytics market. However, we differentiate ourselves by utilising cutting-edge technology to collect and enhance data from multiple sources. Currently, the competition in this field is limited, and we are primarily competing with other startups rather than prominent established players.
Team: Defog.ai is based in Singapore and co-founded by Medha Basu. Apart from us, we have four part-time consultants. We plan to convert them to full-time positions and hire more people within the next few months.
YC- What advice or suggestions would you like to share with other startup founders who are just starting their entrepreneurial journey?
1. Avoid making premature optimisations, explore a wide range of options and be patient. Developing an intuition about the market takes time.
2. Stay current on emerging technologies, and explore how they can be used to your advantage. This could involve striking a balance between exploration and exploitation – training an AI agent to explore the world and then applying its learnings.
3. Maximize the chances of attracting potential customers, employees and investors by creating and sharing valuable content and promoting it to the broadest possible audience. “Publishing a lot and putting yourself out there” can help increase the surface area of luck.
YC- How often should one post content to engage the audience and drive growth?
Rishabh– The frequency of content posting is a balancing act. It’s more of an art than a science. Some individuals might frequently post, for example, 100 tweets per day, while others might publish a blog post every few weeks. There is no one-size-fits-all approach, and it’s essential to experiment and see what resonates best with your audience.
The most important thing is to start and put yourself out there as much as possible. Don’t let the fear of not having a perfect plan prevent you from taking action and experimenting with new strategies. It’s essential to be consistent and post content regularly to engage your audience and drive growth.
YC- Could you share your process for acquiring your initial customers and what a founder should keep in mind when dealing with client customisation requests?
Rishabh- I’ll address the second half of your question. Learn to assert yourself as a novice. While building something valuable, saying no to a customer is okay if their request doesn’t align with your vision. Be confident in upholding your vision.
However, if a customisation request is relevant to the rest of your client base and has the potential to expand your market, it may be worth considering. Remember that customers are not likely to run away simply because you said no to their request. Be clear and respectful in your communication, and focus on the long-term success of your product.
To answer the first part of the question, acquiring initial customers can involve putting yourself out there and increasing awareness of your product or service.
You can optimise for Google ads or SEO or use more relationship-based strategies such as attending events, mentoring others, or speaking at conferences. It’s crucial to find out what works best for you and to strategise accordingly.
YC- How do you determine the pricing for your product, mainly when it’s initially offered for free? When is the appropriate time to start charging, and what should the cost be?
Rishabh- The approach to pricing a product or service is highly dependent on the unique characteristics of the business. There is no one-size-fits-all solution for pricing.
For example, if you are building a marketplace that will benefit from strong network effects, you may need to subsidise both sides to get enough users on board.
For instance, if you are developing a marketplace that will benefit from powerful network effects, it may be necessary to subsidise both sides to attract sufficient users.
On the other hand, if you are a SaaS company with no network effects, charging users from the outset may be more appropriate. In this scenario, deciding on a suitable pricing model, such as usage-based, team-based, or seat-based pricing, is crucial. However, if collaboration is a vital aspect of your product, seat-based pricing may not be the best approach as it could discourage users from adding colleagues to the product to save costs. Instead, alternative pricing models that align with your business and value proposition should be considered.
When determining the price of your product, there are a few options to consider. You can offer access to some features free and charge for others or allow for a certain number of free usage before requiring payment. Alternatively, you can charge based on the number of seats that have access to your tool.
If you opt for a usage-based or seat-based pricing model, your revenue from existing customers will increase as they use your product more. However, if you choose a feature-gated pricing model, it’s essential to be mindful not to increase the payment for existing customers.
Companies like Recipe and Stripe offer the option to retain the original pricing plan for old customers. It’s acceptable to increase the price for new customers, but it’s essential to consider the needs and preferences of your early adopters.
It’s essential to consider the impact of pricing changes on your existing customer base and its long-term growth potential. To ensure customer satisfaction, find a balance that allows your business to grow while maintaining loyalty from your early customers.
One strategy to consider is sending a message to your early users, letting them know that you are increasing the price for future users. Still, they can continue to use the product at its current pricing as a thank-you for their loyalty.
YC- If allowed to go back in time, what areas would you like to focus on differently?
Rishabh- Raising venture capital was an option that I considered when running my company, Loki. We received multiple offers, but at the time, I was firmly committed to being a bootstrapped entrepreneur and valued the freedom that came with it. In retrospect, this decision was a mistake as raising funds would have allowed me to expand the business, hire more staff, and grow aggressively, which could have led to more growth and potentially less stress.
Networking is an area that I wish I had given more attention to in the early stages of my business. Initially, I was hesitant to connect with people in person and preferred communicating online. However, I now realise that the connections and relationships developed through face-to-face interaction are much more powerful and valuable than those formed through online communication alone. This could have led to valuable mentorship, advice, resources, and collaborations that could have benefited the business.
YC- What books do you recommend to other entrepreneurs or startup founders?
Rishabh- There are a few books that I highly recommend to startup founders.
- Zero to One, by Peter Thiel, is an obvious choice.
- Amp It Up by Frank Slootman, the CEO of Snowflake Computing, which focuses on managing your sales team effectively.
- Poor Charlie’s Almanack– It’s not specifically about startups, but the mental models discussed in the book can be applied in many different contexts.
YC- What motivates you to keep going during challenging times? 🎯🎯
Rishabh- For me, I must find joy and fulfilment in the work I do. I have always been passionate about creating new products and experimenting with AI tools. If I were not able to do this kind of work, I would be uncertain about what my career path would be.
YC- What are your career aspirations for when you reach the age of 40?
Rishabh- From my experience, I have discovered that it’s difficult to make solid long-term plans that exceed five years. However, I am dedicated to developing solutions that save people time and improve their lives. I am genuinely passionate about this work and cannot imagine pursuing a different path.
YC- What advice do you give founders seeking acceptance into the Y Combinator program?
Rishabh- We were recently accepted into the YCombinator program. From our experience, some key takeaways are:
- Move quickly through iterations, and don’t delay shipping while waiting for everything to be perfect.
- Embrace the 90/10 rule, where releasing 10% of the product can provide 90% of its value to users.
- Regularly ship updates and actively engage with customers to gather feedback and improve the product.
Rishabh, it was a great conversation. Thank you for your time. Yellow Chapter wishes you the best in your future endeavours.