Process and Panelists

There is screening round at the beginning which is resume based and does not require candidate presence. After that we would have three more rounds.

Round 1

Description of the round

Reasoning and Applications Skill Assessment. In this round we will spend at around 20-30 minutes with the candidate. There is less of a conversation and more of question answering.

What to look out for

The knowledge of the candidate in standard software - Jira and Slack. The knowledge of coding principles and practices.

Sample questions

• How do you write list comprehension in Python?
• How do you assign and monitor tasks in Jira?

Panelists

The panelist for this round will be Pavan Kumar, Lead SDE.

Round 2

Description of the round

​Technical Assessment round. In this round we will spend somewhere around 40-60 minutes with the candidate. The conversation will be focused around the tech-stack mentioned in the projects and work-experience.

What to look out for

Here we will be looking to find out if the candidate has relevant skill-set in the following areas - Database Management, Querying, ML Modeling, Cloud Architecture and General Knack for Problem Solving.

Sample questions

• How does one go about in deciding whether to use Relational or Non-Relational Database in a project?
• If you have a badly performing query, what steps will you take to optimize its performance (make any assumptions)?
• (Pick a project where they have used ML Modeling) Which metrics did you use to evaluate the performance of the model in this project?
• (Pick a project where they have the potential to be deployed on cloud) If you were to migrate this project to cloud (Any of AWS/GCP/Azure) which services would you use and why?

Panelists

The panelist for this round will be Manik Sharma, the CTO of the Org.

Round 3

Description of the round

Soft-Skill Assessment round. In this round we will spend around 60-90 minutes with the candidate. The first part of conversation will take place in a different style - User Research, where the candidate will be doing a user call , the user being interviewer.

What to look out for

​Here we will be looking to find out if the candidate has relevant skill-set in the following areas - User Research, Composure when user turns hostile, converting user requirement to features in the platform.

Sample questions

The interviewer will layout their problems with the current setup. If the candidate tries to solve the user problem on the call the interviewer will become un-reasonable and hostile and will try to drive the conversation in pointless circles.

Panelists

The panelist for this round will be Roshan Raj, the CEO of the Org.

Feedback + Rating Mechanism

Rating Mechanism

The candidate is scored separately out of 10 in different skill areas and an average of this rating is calculated - the skill rating. This is not all, there is second rating called Overall Rating which is out of 4. The second rating captures the gut feeling as well as forces the interviewer to take clear stance - positive or negative.

​Round 1

Skill Areas Tested

Application Competency and Coding Practices.

Guideline for Comments and Feedback

• In case the candidate is not familiar with the tech-stack they are tested on but have good skills in adjacent tech-stack, a note should be made.
• A note should be made if the candidate blazes past the questions with a very high pace.
• A note should be made if the interviewer feels there is a use of unfair means (usually LLMs) going on.

Round 2

Skills Tested

​Database Management, Querying, ML Modeling and Cloud Architecture

Guideline for Comments and Feedback

• In case the answer for Relational vs. Non-Relational is satisfactory. Give them a toy schema to design and note down all the things they considered in making the schema.
• The question asks the candidate to make assumptions about the schema to answer the 'Optimize query speed', note down these assumptions
• In case the candidate is able to hash out why they used a particular evaluation metric in the ML Modeling project, ask them the mathematical intuition behind the metric and note down the response.
• In case the candidate is able to form a sound toy cloud architecture, ask them to scale this architecture and note down the response.

Round 3

Skills Tested

​Problem Solving, User Research Ability, Stress Management, Kanban

Guideline for Comments and Feedback

Make extensive notes throughout this round

• Note down how long does the candidate speak vs. listen for each particular pain point. While listening is important, it is also important to guide and advance the conversation.
• During the user the call make note on how the user handles the hostile situation. One good way is too see if they 'Yes and'
• Post the user call, make note how the candidate goes about grouping and creating hierarchies of the problems shared during call
• Take a screen shot of the Kanban made. The Kanban is to test how they will communicate the findings to different stakeholders.

Decision on candidates

• The first round is an elimination round, any candidate with a score of 1 and 2 is not taken into further rounds.
• Second and Third round are deep-dive rounds. We follow the following criteria:
• • All 4's : HIRE
  • One 4, Rest 3 : INTERNAL DISCUSSION
  • All 2's and 3's : ONE MORE ROUND (The low scoring round)
  • Even a single 1 : REJECT

​

​