In the past year, I’ve been deeply involved with the HRTech industry, a $400Bn industry globally which is fast evolving to adopt AI to reinvent the recruitment space. As the job market shifts towards being candidate-driven (from being employer-driven) and as candidates/ employers shift towards tailored online job platforms (from traditional recruitment firms), it’s an exciting time to be in the HRTech space, an industry ripe for disruption. I’d like to share an interesting case-study from the war-trenches at Indeed (Largest job-search engine in the world; 250Mn MAU) around how we adopted AI to disrupt employer-jobseeker matching at scale. I’d also like to talk about how we managed to strike the right balance between responsibility and innovation, which is very important for platforms at our scale. As a seasoned entrepreneur (3x startups, 1x unicorn exit) turned product leader, I’d also like to share some experiences where I had to learn/ unlearn product principles depending on the industry/ stage of the company – around how some of the product principles used in startups/ mid-stage companies was not a good fit for late-stage companies and vice-versa.
HRTech is a $400Bn industry which is fast evolving to adopt AI to reinvent the recruitment space. I’d like to share an interesting case-study from the war-trenches at Indeed (Largest job-search engine in the world; 250Mn MAU) around how we adopted AI in the HR Tech industry to disrupt employer-jobseeker matching at scale.
Key areas of learnings :
1. How AI can empower people to make better Career decisions?
Salary is a critical matchmaking parameter for Career decisions, but both jobseekers & employers are not comfortable sharing it upfront. In the absence of factual salary data, how do we solve the Salary transparency & matching problem between employer-jobseeker at scale using AI based salary-predictions.
2. How to prepare for AI-imperfection (wrong predictions) to avoid blowback?
In spite of large-scale data & continuous feedback loops, AI predictions are far from ‘perfect’ even at scale. How can product teams prepare beforehand for AI-imperfection i.e. when data-prediction-experiments don’t go as planned, to avoid blowback from clients/ enterprises/ users.
3. Using AI responsibly at scale.
Solving problems for our users comes with the added responsibility of accounting for their emotions. Ex: Job search is an intense and stressful time for jobseekers, and a right/ wrong AI-predicted-Salary estimate can make/ break their career decision. Hence, how can product teams run AI experiments responsibly by factoring in consumer psychology on both sides of the marketplace (employer/ jobseeker).