Skip to content
Guruprakash Sivabalan

Founder story

Building Xobin

I’m Guruprakash Sivabalan, Founder and CEO of Xobin. Xobin helps hiring teams see what a candidate can actually do, through skills assessments, work samples and structured interviews. This is how it started and what I have learned about the problem.

The hiring problem

Most hiring decisions rest on a résumé, a conversation and a few references. Each tells you something. None of them shows you how a person handles the actual work. When a decision like that goes wrong, the cost spreads well beyond one salary, into the team and what it can build.

We put it plainly in Why Xobin. Hiring teams work hard. What they need is clear, tested evidence of what someone can do.

Starting with Amrit

In March 2016, my co-founder Amrit Acharya and I set out to talk to 100 CTOs and HR leaders about how they hired experienced engineers. We wanted to understand the problem before building anything. The conversations kept returning to the same point. Engaging, assessing and hiring engineers took far too long.

A year later I wrote up that first year in Happy Birthday Xobin!, including our first customers and what we learned from them. Today Xobin’s founders page introduces Amrit as COO. I’m the Founder and CEO of Xobin.

Work samples and structured interviews

A work sample asks a candidate to do a small piece of the real job. It shows what they understood, what they checked and where they used judgment, which is closer to the job than a description of past experience.

A structured interview asks every candidate the same job-related questions and scores the answers against criteria written in advance. That makes answers comparable and makes it harder for a confident conversation to stand in for evidence.

Hiring teams use these results alongside their own judgment. Xobin gives the people who make the decision clearer evidence.

One research finding

In its report published on 9 September 2026, Xobin tested AI workflows against the 683 skill groups in its assessment framework. In the Human vs AI Skills Report, 72% of those 683 skill groups met the criteria for full or partial AI delegation under at least one tested setup. Each percentage in the study describes a tested skill group within this benchmark.

For hiring, it changes the question. As I told HRD America, the useful question is which parts of a task a candidate must do independently, and which they need to direct and check.

Sources