Human Stories in AI: Abbas Merchant@Matics Analytics
Watch on YouTube →
Overview
Abbas Merchant, founder and CEO of Matics Analytics, shares his unconventional journey from dropping out of high school to pursue his family's retail business, to returning to complete his education, and finally pivoting to computer science and AI/ML. His company, Matics Analytics, focuses on providing AI-powered marketing solutions, customer retention strategies, and fraud detection for financial institutions, leveraging machine learning models like XGBoost and LightGBM.
Key takeaways
- Abbas Merchant's journey highlights the value of education, even after initial rejection, as a foundation for pivoting careers and entrepreneurship.
- Matics Analytics addresses the gap in AI adoption for SMEs by focusing on practical applications like ML-powered marketing and customer retention.
- The company utilizes advanced ML techniques like XGBoost and LightGBM, combined with explainability tools (Shapley values) and rigorous evaluation (lift/gain charts, A/B testing), to deliver tangible business value.
- Abbas emphasizes a 'calculated risk' approach to entrepreneurship, informed by prior experience and market research, contrasting with earlier impulsive decisions.
- Continuous learning is a core philosophy, driving both personal growth and the company's ability to adapt and innovate in the AI space.
- The iterative process of data analysis, involving feature engineering, modeling, evaluation, and explainability, is critical for refining ML solutions.
Chapters
- Abbas Merchant is the founder and CEO of Matics Analytics.
- Matics Analytics uses AI and analytics to transform enterprise data into intelligent actions.
- The series 'Human Stories in AI' features career journeys of AI experts.
- Abbas Merchant considered dropping out of 11th grade to join his family's retail and distribution business for consumer electronics.
- His parents rejected this idea, emphasizing the importance of education.
- He initially believed practical business experience was more valuable than academic knowledge.
- After convincing his parents, Abbas worked in the family business for three years, experiencing good revenues and profits.
- He felt a lack of growth due to perceived lack of qualifications, especially with the booming e-commerce market in India (Amazon, Flipkart).
- He recognized a 'missing piece' and decided to switch gears to complete his schooling.
- With only four months left before final board exams (12th grade), Abbas studied 13-14 hours daily.
- He successfully cleared his exams with good grades, supported by family and friends.
- He explored business courses (BBA, MBA) but felt they wouldn't help him address the high competition from e-commerce.
- Abbas questioned how his schooling, particularly statistics, could help his career path.
- He explored the tech industry's use of statistics, driven by the e-commerce challenge.
- He decided to pursue computer science as his career.
- During his computer science studies, he found college courses lacked practical application of statistics.
- He discovered Josh Starmer's StatQuest videos on YouTube, explaining practical implementations of stats in AI/ML and data science.
- Within six months of learning ML, stats, and Python, he secured an AI/ML R&D intern offer.
- After his internship and subsequent job, Abbas worked in the AI/ML field for over five and a half years across various industries.
- He enjoyed continuous learning but felt a desire for change beyond climbing the corporate ladder.
- He observed a gap in AI/data analytics adoption among Small and Medium Enterprises (SMEs) compared to Fortune 500 companies.
- Abbas spent 1.5 years conducting market research alongside his job, speaking with business owners and industry experts.
- He identified three key issues for SMEs: lack of awareness about data value, affordability concerns, and uncertainty on how to utilize AI.
- The rise of tools like ChatGPT increased general AI awareness, highlighting the opportunity for Matics Analytics.
- Six months prior to the interview, Abbas incorporated Matics Analytics.
- He quit his senior lead data scientist position at Moody's Analytics to go 'full in' on his company.
- This entrepreneurial step was a 'calculated risk,' unlike earlier 'jumping off a cliff' decisions, supported by savings from working from home during COVID-19 and family financial stability.
- Matics Analytics is currently working with clients on ML-powered marketing and customer retention use cases.
- They also handle fraud detection in the financial domain.
- A key focus is building machine learning models for propensity modeling, expected spend prediction, and channel preference identification.
- The core problem is selecting the right target audience for marketing campaigns to reduce costs and lost opportunities.
- Matics Analytics builds multiple ML models, including propensity models (predicting likelihood of opening a credit card), expected spend models, and channel preference models.
- They utilize historical transaction, product, and campaign data, employing techniques like XGBoost and LightGBM for tabular data.
Summary, takeaways, and chapters were generated by AI from the video's transcript and may contain errors. The video belongs to its creator, StatQuest with Josh Starmer.