Me postulé en línea. El proceso tomó 5 semanas. Acudí a una entrevista en Fractal (Bengaluru) en mar 2021
Entrevista
It was a systematic interview process:
Round 1: DoSelect coding round - All relevant questions - including AI/ML-based MCQs and ML Modelling-based questions. The time constraint was the hardest part of this round, otherwise, questions were of average difficulty if you know the fundamentals.
Round 2: Technical (Senior Data Scientist) - Starting from project experience, it covered major important ML/DL concepts and questions from the skillset mentioned in the Resume.
Round 3: Technical + Managerial (Lend Data Scientist) - It was somewhat similar to the previous round, more focused on the Techanalities of my project in detail. Why? and Why not? questions from my previous work and the use of various ML/DL concepts and their alternatives. The interviewer was also interested in knowing my role and responsibilities in the current organization.
Round 4: Technical + Managerial (VP) - It was focused more on the kind of projects I worked on and my contributions (in detail), career objective, and aspirations, it also delves into my skillset and target in terms of learning.
Round 5: Managerial (Manager) - The interviewer was interested in knowing the vastness of all my previous projects and my role in those. I was being asked about my understanding of various Data Science concepts in detail (i.e. NLP, Timeseries Analysis, Mathematical optimization, etc.). I was asked a basic program logic even in this round.
Round 6: HR: Duration was ~30 mins, and I was asked about reasons behind job change and aspiration. HR asked for depth in my answers which eventually helped her getting many other questions answered.
I felt all rounds were elimination rounds, including the final HR round. During this entire process, the support and details provided by my TA contact person Jesvin were really helpful. It made me create a rough plan and decide on the depth of my answers.
I received a positive response and an offer letter in the next 4 days. It was challenging and an amazing experience.
Preguntas de entrevista [1]
Pregunta 1
Problem-solving.
In-depth questions from previous projects and skill set.
A strong grasp of fundamentals is necessary.
Why you used certain techniques and algorithms over others?
Client round was more tough. Screening also was around Reinforcement learning. Which I happened to pass. But cannot pass the client round second time. The interview process was around Artificial Intelligence and RAG.
Me postulé a través de una recomendación de un empleado. El proceso tomó 1 semana. Acudí a una entrevista en Fractal (Nueva Delhi) en sep 2025
Entrevista
The whole process took almost a week, including 2 Technical interviews, 1 HR round and offer release. Many thanks to Jesvin for making all this possible in such a short period of time.
Coming to the interview
1st round started with basic Python questions and data manipulation, followed by questions on previous role, projects, agentic frameworks, RAG, and retrievers.
2nd round was more in-depth, with a focus on details about Agent implementations, MCPs, Evaluation and improvement techniques.
3rd round (HR) was much simpler, with questions related to previous work experience.
Overall, the process was quick, and every interviewer was supportive and gave feedback properly.
Preguntas de entrevista [1]
Pregunta 1
Difference between RAG and Agentic RAG
How to evaluate and improve the performance of a deployed model/agent
The interview process typically begins with resume screening, followed by an initial HR call to discuss background and expectations. This is succeeded by technical assessments or assignments to evaluate role-specific skills. Shortlisted candidates then proceed to one or more rounds of technical or panel interviews, where problem-solving, communication, and cultural fit are assessed. The process concludes with a final discussion, reference checks if required, and offer rollout to the selected candidate.