Cliently is the first entirely artificial intelligence-powered sales engagement application in the world. It automatically develops personalised real-time AI forecasts that inform salespeople which accounts and contacts to engage with and which actions to take to increase sales and save countless hours.
AIM: From data scientist to data science consultant to digital marketing consultant to a chief operating officer – you have a diverse professional path. Which position is your favourite? How did you compose your career path?
Harsh Gupta: In startups, you have to wear multiple hats. I always wanted to have my own company, for that, it’s very important that you have to be multidimensional. I have been naturally good with marketing and data science, and if you think about it, data science is nothing but using stats in marketing. I like to be in charge of many departments or at least be involved, so the position of COO or CEO fits me most because then I am involved in the growth of the company, overlook operations, and make strategies. My actions have an immediate impact, and that is what I love, more challenges.
I started my career as a data analyst, moved to a data scientist with a good aptitude, deep knowledge and masters. However, I don’t recommend masters for everyone. I started getting consulting work along with my full-time job because of my speaking gigs in conferences. I prepared content and showcased it to an audience, who would later speak to me and take my consultation.
In 2020, I started my own company, which was acquired by Cliently later. I had good interests from investors, too, and that was all because of my marketing efforts and expertise.
I now work as COO and Head of AI at Cliently; we are working with 150+ businesses and giving them AI predictions in a very convenient manner for their sales teams.
AIM: What, in your opinion, are the hurdles associated with building a data science team?
Harsh Gupta: Finding right talent. I have taken 1000+ interviews and rarely find less than 1% hireable. And this filter is purely based on knowledge and coding skills. When I start testing out aptitude, the right fit, etc., I lose a lot of candidates. So, I have a very limited pool to select. Further, I lose many candidates to competition, since we are not a huge company.
Another hurdle is that we have to mould them in a certain way of thinking, which data scientists are not trained in. Business first – what business outcome am I getting from the projects. The goal is never to build clean data or build models, but giving companies action plans, recommendations, and a strategy roadmap is equally important. This is where data scientists have to think a lot and speak the language of business. So, I have to make the team start thinking like this.
AIM: It’s great to know that you were inspired by the power of data science in marketing and sales. Which incident inspired you?
Harsh Gupta: Spend some time with your marketing and sales friends/colleagues, and soon you will realise there is a huge gap between data science and what they do. Data science would empower them so much, but they don’t know what to do with data, and communication with data scientists is bad (if they have a team of scientists).
AIM: What are the habits required to become a successful data scientist, and what is your dream data science project that you aspire to complete?
Harsh Gupta: Live and dream in data. Think of all possible outcomes, be creative about formulating problems. Problems can be solved in so many ways. You will rarely get a problem where they tell you that this is the target and these are the input variables. You will have to be talking to multiple stakeholders, convincing them/showing the value of a project, then getting datasets from many places, coming up with many hypotheses, doing a lot of preliminary investigation, creating subsets of datasets, taking different targets. All this requires immense thinking, creativity and awareness on your part. A good way to inculcate this habit is by reading a lot – WSJ, Bloomberg, Economic Times, analytics blogs and magazines, watching CNBC, attending conferences and webinars. You will see people discussing problems and finding solutions – folks discussing best practices, challenges that will make you think.
My dream project would be building a recommendation system as complex as that of Netflix, Youtube, Tik Tok – so precise.
AIM: It’s exciting to know that you have additional expertise developing and leading AI initiatives across various industries. Which industry was challenging to you?
Harsh Gupta: Non-profits, management doesn’t see much value, and you have to convince so many people to bring a project to life. And very often, a project won’t see the light of the day because of too many apprehensions.
On the other hand, I find banking challenging because of the nature and amount of data – tonnes and tonnes of transactional data. And the industry is so mature that they are always looking for the highest accuracy and mature applications of data science.
AIM: Do you believe that a data scientist should have a great understanding of marketing? What is your opinion?
Harsh Gupta: Yes, unless you are in a very specific niche like Computer Vision, NLP, etc. Most data scientists work with structured data to give insights, predictions and recommendations to companies. And 80% of the time, it would involve marketing, sales.
Doing customer churn analysis? Finding out clv? Understanding the sentiment of customers? Lead Scoring? All ties up to marketing and sales in one way or another.
It will be hard to understand their requirements if you don’t understand marketing or speak their lingo. You have to see through their lens to get the gist of their problems.
You will be hearing a lot of these words in your day to day work –
ClickThrough Rate (CTR)
Ideal Customer Profile (ICP)
Key Performance Indicator (KPI)
Pay Per Click (PPC)
Return On Investment (ROI)
Top of the Funnel (TOFU)
AIM: What are your favourite data science/ AI books?
Harsh Gupta: I never read a book; I read a lot of magazines, newspapers. Always checking out new GitHub libraries, attending conferences – that is how I keep myself updated.
AIM: What advice would you provide to entrepreneurs and those considering a career in data science?
Harsh Gupta: Entrepreneurs – find co-founders who can compliment you; if you have a data science background, you very likely need somebody great with marketing, someone with enterprise sales experience, somebody with web development, etc. Unless you have that solid team of founders, it will be tough, as today your business is always somehow competing against giants like Amazon, Facebook, Microsoft, etc.
Those considering a career in data science – professional life is very different from academic. You will navigate through many challenges, so be absolutely sure where you want to work—for example, some people like working on products, whereas some like consulting. Be very clear with machine learning concepts and have excellent coding skills, at the very least in Python and SQL. This will make interviews much easier.
Subscribe to our NewsletterGet the latest updates and relevant offers by sharing your email.
Nivash holds a doctorate in information technology and has been a research associate at a university and a development engineer in the IT industry. Data science and machine learning excite him.