Skip to content
dronacharya.ai

AIRA Matrix

PlacementData and ML

Built from 1 candidate report with 1 first-hand accounts, covering 2025-26.

Hiring process

Roles: AI/ML Engineer Trainee

2025-26 · Placement

  1. 1

    Resume shortlist

    Direct resume shortlisting
    eliminatory

    relevant projects

    No coding or aptitude test; candidates picked for technical interviews on resume, with focus on relevant projects.

  2. 2

    Technical interview

    Technical round 1
    Google Meet60 minvirtual

    thesis and projects · loss functions · batch normalization train vs inference

    Rapid deep dive into fundamentals. The document labels round 1 as projects and applications but lists basic ML/DL questions.

  3. 3

    Technical interview

    Technical round 2
    Google Meet60 minvirtual

    activation functions · sigmoid vs softmax · class imbalance in segmentation · optimizers · transformer attention · positional encoding · PyTorch or TensorFlow syntax

    Deep learning fundamentals plus project and application focus; questions were open ended.

  4. 4

    HR interview

    HR / management round
    Google Meet20 minvirtual

    behavioral scenarios · unclear requirements · failure and feedback · teamwork · why this company · strengths and weaknesses · questions for HR

    Roughly sixteen behavioral questions listed.

Advice from candidates

7 pieces of advice, in their own words. Names removed.

  • Be sure to have thoughtful questions prepared for the interviewer. This is one of the most important parts of the HR round. Use the pre-placement talk notes to ask smart questions about the training program, project types, or the "innovation cycle."

    HR / management round · 2025-26

  • While I had a good grasp of the concepts, I could have improved by practicing articulating the mathematical intuition behind them more crisply. For example, practicing whiteboard explanations of backpropagation through a B-Norm layer or the attention mechanism.

    interview · 2025-26

  • Focus on Projects: Have at least 1-2 strong computer vision or deep learning projects ready to discuss in-depth. Projects on medical images are a significant plus.

    preparation · 2025-26

  • Know Your "Why": Be ready to explain your problem-solving approach, why you chose a specific architecture or loss function, and how you would handle limited guidance.

    interview · 2025-26

Loading the other 3…