As the CEO and co-founder of Findem, Hari is responsible for driving the company’s overall direction and strategic growth, as well as overseeing its day-to-day operations. He’s a serial entrepreneur and accomplished technologist, with nearly two decades of experience building companies and creating trailblazing technology solutions.
Hari was previously the co-founder and CTO of Instart, where he led the company’s technical vision and translated customer requirements into realizable, innovative solutions. During his time at Instart, he co-authored more than 50 patents.
Hari has also held senior-level engineering positions with Aster Data, where he worked on all features across the entire development stack, as well as with the Solaris Cluster group at Sun, where he contributed critical modules of software.
You have been a successful entrepreneur having successfully launched two startups. Could you discuss the eureka moment at your first startup Instart, when you realized that scaling a team is a major problem for most entrepreneurs?
It wasn’t just one, but more a combination of a couple different experiences. We reached a point at Instart where we were on an extremely fast growth path, including expanding the company internationally, and that presented a special set of challenges. Now, we’re trying to build an exceptional team that’s truly diverse, and doing so in short order and across continental borders. As we were competing with other startups for talent and rushing to scale our team, we ended up making a couple of bad hires, which set us back and created a lot of frustration. Other bumps in the road came when we tried to communicate our employee wish list to recruiters. The process was highly error prone, and we found ourselves compromising many times on the right hire in the spirit of closing quickly. These were hard-hitting lessons and ones that challenge nearly all entrepreneurs, but I’m thankful that they sparked the idea and fed the fire that led to Findem.
Could you then discuss the genesis story of launching Findem?
Findem was really a direct result of the mistakes I made in hiring and scaling earlier in my career. As any entrepreneur will tell you, building exceptional teams is the single most important factor in a business’s success. It’s also exceedingly difficult. As someone with an engineering background, I’m drawn to solve some of the hardest problems that lead to the biggest impacts, and I was motivated by this particular challenge. Finding the right hires who can seamlessly gel with the culture of the company and have the competencies required to discharge the job is much harder than it sounds.
Traditionally, the only way to crack the talent scaling problem was through brute-force, along with a human element—and the process was littered with errors, bias and inefficiencies. As I examined it further, it struck me that it’s actually a data problem at its core, and the correct way to solve it is to approach it like a data problem. Using AI and deep analytics, we’ve brought a successful new approach to the process by enabling HR leaders to search for candidates based on desired attributes rather than by keywords or titles on resumes. Companies being drawn toward data-based recruiting because it’s more efficient, reduces costs, improves equity and results in better quality hires. Findem started as a passion project and now we’re thriving, particularly among enterprises that encounter more hiring aches, pains and expenses than their smaller counterparts.
How important is data when it comes to hiring?
Data is crucially important when it comes to making effective hiring decisions. For example, when companies are trying to build more diverse teams, tracking employee and candidate data is often an afterthought. However, it’s vital that diversity, equity and inclusion (DE&I) initiatives start with transparency about the current, data-informed state of the organization—analytics can show you everything from the diversity of your leadership, to how you’ve been tracking on diversity over the past five years, to compensation discrepancies, to turnover rates of diverse employees. It’s important to note that data tracking should extend not just to gender and race, but other factors as well, such as age, religion, disability and military service. When you have that data, you can start mapping out your goals and truly work toward a diverse and inclusive culture.
Also, when it comes to building that diverse and inclusive culture through hiring, it is very important to monitor the talent pipeline to ensure you’re nurturing diversity right from the start of the candidate search. This is impossible without the right data.
Pipeline analytics are also key to understand what’s working or not in your diversity recruitment efforts. How quickly are diverse candidates recruiting? Which recruiters are really moving the needle when it comes to filling the pipeline with diverse candidates? Are you sourcing from geographic areas where there are a higher percentage of diverse candidates? Data can provide the answers to all of these questions that you’d be unable to answer otherwise.
Data is also at the core of predictive analytics, where historical data is used to unearth talent who will excel within your company. Predictive analytics can tell you how likely a candidate to perform well in a certain role, their risk for turnover, if they’ll be successful in a remote position and other information that can help you pinpoint candidates who are most likely to flourish.
What are some of the data sources that Findem collects information from?
Findem aggregates all publicly available people data, which is verified and triangulated across multiple sources, for the purpose of recording and learning about a potential candidate’s attributes. We have an library of more than 1 million attributes for every individual. We can enrich this data and discover new attributes if our customers choose to integrate their internal HR tools with Findem. Some examples of the public information we aggregate include Census data, product category information data, company financial data, market data, patents and publications data, educational data, and productivity and skills data.
How can employers best use the Findem platform to match with the ideal candidate?
To match with ideal candidates—whether they’re active or passive—employers can use our platform to search for them based on a combination of over 1 million attributes. Attributes can be tangible, such as whether someone is female, a previous founder or worked for a top-10 VC-funded startup, as well as intangible, such as whether someone embodies the company’s values, possesses an entrepreneurial spirit or is a go-getter. These attributes provide a data-informed picture of each individual and can be used to find the exact fit to fill an open position.
Attributes can be matched across internal employees, ATS profiles that are enriched with the most up-to-date information and external candidates. Typically, companies start with an ideal candidate profile and build a talent pool of every person who matches that ideal candidate’s attributes, although some opt to build an attribute search from scratch.
Another unique approach they can take is to analyze the attributes of someone who is a superstar employee—they could be inside or outside the company that’s hiring—and then architect a search for candidates who are essentially their clones, meaning they also possess those exact attributes. Say they know someone who excels at remote work, is loyal and was a CMO at company that was successfully acquired, an employer can simply search on our platform for a set of copycat individuals.
How does Findem avoid unintentional gender or ethnic bias from its machine learning process?
The inadvertent bias that gets introduced without any visibility into talent distribution—AKA diversity—when picking a particular location or attribute to search for is a source of unconscious bias. Findem provides an aggregated summary of the talent distribution dynamically by location and various search attributes and gives this visibility to the people team.
We also diminish these biases through attributes-based searches that can be done without human involvement, by obfuscating candidate PII information when conducting manual reviews and by automatically adding weights to the pipeline to ensure it’s as diverse as possible.
One interesting concept is how Findem enables employers to find new attributes for talent searches. How does this process work?
Findem enables new attributes to be uncovered in a number of ways. One is by looking at other companies and the people they have hired at different times. For example, if a company is planning to raise a Series B round or go public, it may want to understand how companies that were very successful in similar endeavors were staffed. Our platform lets employers see the attributes of those people and use them in their own talent searches.
Similarly, you can do this with your own superstar employees and internal systems. It’s possible to use your internal human resource information system (HRIS) to distinguish your top performers, and then you can identify attributes that are common to them and use that to feed future searches.
Is there anything else that you would like to share about Findem?
One of our biggest areas of focus right now is fulfilling our vision of making our talent sourcing solution completely self-service. A day-one goal for us was to build a platform that was simple enough for anyone within the HR function to use, and we’re making huge strides right now in reaching that milestone.
Thank you for the great interview, readers who wish to learn more should visit Findem.
- Neural Networks Learn Better by Mimicking Human Sleep Patterns
- 7 “Best” AI Translation Software & Tools (November 2022)
- Conversational AI Is Making Customers and Employees Happier. Here’s How.
- Electricity Helps Find Materials That Can “Learn”
- Lack of Trustworthy AI Can Stunt Innovation and Business Value