Data Science Tutor Job — Remote, Freelance, Rs 500-1,500/hr

Role	Online Data Science Tutor (Freelance)
Pay	Rs 500 – Rs 1,500 per hour
Type	Freelance, part-time, work from home
Location	Remote. India-based tutors preferred; global applicants welcome
Hours	Flexible, mainly 5 PM – 9 AM IST
Students	Mostly USA, Gulf, Europe, Australia
Apply via	MEB tutoring jobs hub

The Data Science tutor job at MEB involves running 1:1 live online sessions and providing homework guidance within those sessions, mainly for students in the USA and the Gulf. Students who request this role are typically enrolled in graduate-level courses, professional certification programmes (such as those from Coursera or edX), or undergraduate data science and statistics degrees at US universities. Sessions frequently require hands-on work in Python or R — writing, debugging, and explaining code directly on a shared digital whiteboard — alongside conceptual explanations of statistical theory and model behaviour. A pen tablet is not optional here; freehand annotation of equations, decision boundaries, and confusion matrices is a routine part of the session.

What the role involves

Running live 1:1 sessions in which you explain data science concepts, walk through statistical methods, and debug code alongside the student on a shared whiteboard.
Guiding students through their own problem sets and projects — explaining the method, not supplying the answer.
Working through machine learning model implementation, interpretation of results, and evaluation metrics in real time, under deadline.
Helping students understand Python or R libraries (NumPy, pandas, scikit-learn, ggplot2, tidyverse) through worked examples built live in the session.
Adapting explanations instantly when a student’s conceptual gap turns out to be deeper than the presenting question — which in data science is more often the rule than the exception.

Topics you will be expected to teach

Exploratory data analysis (EDA) and data wrangling with pandas and NumPy
Probability theory and statistical inference (hypothesis testing, confidence intervals, p-values)
Linear regression, logistic regression, and generalised linear models
Decision trees, random forests, and ensemble methods
Support vector machines and kernel methods
Unsupervised learning: k-means clustering, hierarchical clustering, PCA and dimensionality reduction
Neural networks and introductory deep learning (architectures, backpropagation, optimisers)
Natural language processing fundamentals (tokenisation, TF-IDF, word embeddings)
Model evaluation: cross-validation, ROC-AUC, precision-recall, bias-variance trade-off
Feature engineering and feature selection techniques
SQL for data retrieval and aggregation in a data science workflow
Data visualisation principles and implementation in Matplotlib, Seaborn, or ggplot2
Bayesian inference and introductory probabilistic modelling
Time series analysis: stationarity, ARIMA, forecasting evaluation

A problem you should be able to solve

You are given a binary classification dataset with 10,000 observations and 25 features. After training a logistic regression model, you obtain an accuracy of 92% on the test set, but the recall for the positive class is only 18%. Explain precisely why this discrepancy occurs and describe the steps you would take — including any changes to the decision threshold, class weighting, or resampling strategy — to improve recall without discarding logistic regression entirely.

If you cannot set this up and solve it in under five minutes without looking anything up, this role is not the right fit.

Who we are looking for

Subject mastery

Data science draws on probability, linear algebra, statistics, and programming simultaneously. Mastery here means you can derive why gradient descent converges under certain conditions, explain what the covariance matrix represents geometrically, and write a clean cross-validation loop in Python without referencing documentation — all in the same session. Knowing how to run scikit-learn functions is not sufficient. Students will ask you why the algorithm behaves the way it does, and your answer must be correct and immediate.

Speed and accuracy under deadline

Data science sessions at MEB often arrive with tight turnarounds — a student working through a graded assignment due the following morning needs a tutor who can identify the conceptual error in their model pipeline, explain it clearly, and guide a correction in a single session. You must be able to read unfamiliar datasets, spot data quality issues, and reason about model choices quickly. Slow or approximate answers in this subject cause compounding confusion that a single session cannot repair.

Education and background

A postgraduate degree in statistics, computer science, mathematics, electrical engineering, or a directly related quantitative field from IIT, IISc, ISI, NIT, or an institution of equivalent standing is the standard we apply. Candidates from other institutions are considered where they can demonstrate exceptional applied data science experience — for example, a strong record of published work, verified industry projects involving real modelling decisions, or a sustained history of graduate-level tutoring in this subject specifically.

Setup, availability and communication

You need a reliable laptop, stable broadband, a working camera and microphone, and a pen tablet. Sessions are conducted on a shared digital whiteboard; typing equations or code descriptions into a chat box is not acceptable. Most of MEB’s students are in the USA and the Gulf, so the bulk of work falls between 5 PM and 9 AM IST. Your English must be clear and precise — not merely functional. Students are non-Indian and expect explanations they can follow without difficulty.

Do not apply if

You need a guaranteed monthly income or a minimum number of hours each week.
You cannot work reliably between 5 PM and 9 AM IST on weekdays or weekends.
You do not own a pen tablet — this is a hard requirement, not a preference.
Your data science knowledge is limited to running library functions without understanding the mathematics behind them.
You expect to look up method details, formula derivations, or syntax during a live session.

What this job is not

This is not salaried employment. There is no fixed monthly pay, no retainer, and no guarantee that work will arrive in any given week. The volume of sessions offered depends entirely on what students request, and that fluctuates. This role is not a way to complete students’ graded assignments or projects on their behalf; tutors guide students to understand and solve problems themselves, and that boundary is enforced strictly. This is also not a fixed-shift position — you choose which offered sessions to accept, but you are expected to be dependable when you do accept one.

Pay and payment terms

The pay rate for this role is Rs 500 – Rs 1,500 per hour. The exact rate for any given piece of work depends on the level of the content, the complexity of the session, the deadline involved, and the specific task assigned. The fee is agreed before the work starts. You may accept or decline any assignment without penalty. Payment is made on time. There is no probationary pay cut and no platform fee deducted from your rate.

How work is assigned at MEB

Work is offered job-by-job. When a student request comes in that matches your subject profile, it is offered to eligible tutors on the platform. You choose whether to accept it. MEB distributes work fairly among verified tutors rather than routing everything to a small group. There are no guaranteed hours and no minimum work per month. Freshers are eligible for this role only where subject depth is genuinely exceptional — the data science tutor job involves graduate-level content regularly, and there is no ramp-up period built into the process.

Academic integrity rules for tutors

Tutors at MEB guide students to understand and solve problems themselves. A tutor does not complete graded coursework, take-home exams, or project submissions on a student’s behalf. Explaining a method is the job; doing the work in place of the student is not. Tutors must not share personal contact details with students or negotiate fees directly with them; doing so ends the engagement immediately and permanently. MEB’s full policy is published at myengineeringbuddy.com/trust/academic-integrity/. Read it before applying.

Selection process

Submit the application form on the tutoring jobs hub.
Shortlisting based on subject depth, educational background, and the information provided in the application.
A subject test covering core data science concepts and a short mock session conducted on a shared digital whiteboard using a pen tablet.
Onboarding for successful candidates, after which work is offered job-by-job as student requests come in.

For questions about the process, reach out via WhatsApp +91 8971 383660 or email meb@myengineeringbuddy.com.

Questions from applicants

Do I need to know both Python and R, or is one sufficient?: Most student requests at MEB for the data science tutor job come in with Python-based coursework, so Python fluency is effectively required. R is an advantage for students enrolled in statistics-heavy programmes at certain US universities. If you are strong in Python and have working R knowledge, that covers the large majority of sessions. Tutors who know only R will see a significantly smaller volume of offered work.
Is there a specific machine learning framework I need to be proficient in before applying?: scikit-learn is the library that appears most frequently in student sessions, followed by TensorFlow and PyTorch for students working on deep learning components. The selection test will include live coding questions, so your proficiency needs to be genuine rather than theoretical. Being able to explain what a function does is not the same as being able to use it correctly under time pressure.
I have a strong statistics background but limited industry experience in data science. Am I eligible?: Eligibility is determined by whether you can teach the content to a graduate student accurately and in real time, not by job titles on a CV. Candidates with strong statistics and mathematics foundations who can also code competently in Python are well suited to this role. The subject test will make clear quickly whether the combination holds up in practice.
How many sessions per week should I expect once onboarded?: There is no fixed number. Work depends on what students request and how many tutors are available for a given request at the time it comes in. Some weeks involve several sessions; others may involve none. Tutors who accept assignments reliably and perform well tend to receive more offers over time, but MEB does not guarantee a minimum workload. Plan your finances accordingly before applying.
What does the mock session in the selection process involve?: The mock session replicates a real student interaction. You will be given a data science problem and asked to explain it on a shared digital whiteboard using a pen tablet, as you would in a live session. The assessor will play the role of a student and may ask follow-up questions or introduce a conceptual misunderstanding for you to identify and correct. The goal is to see whether you can teach clearly, accurately, and without preparation time — not whether you can prepare a polished explanation in advance.