Data Scientist III | Pricing, Fraud and Personalisation

Meesho's data science team works on the problems that come with running a marketplace for price sensitive buyers across India: fraud detection, inventory optimisation, demand forecasting and making the platform work in Indian languages. This role builds models and runs experiments against those problems, and the specifics in the posting are commercial rather than abstract. You model seasonal demand to predict organisational metrics, design discount programmes that help resellers sell more, work out reseller preferences to surface the right products, and help resellers read what their end customers actually want. You also mentor junior data scientists. The bar is 4 to 7 years in a fast paced organisation, ideally business to consumer. Apply if you want applied data science with a direct line to revenue, and you are comfortable owning A and B tests end to end rather than handing them over.

Requirements as published:
- Bachelor's or Master's degree in computer science or a similar field.
- 4 to 7 years of experience as a Data Scientist in a fast paced organisation, preferably business to consumer.
- Familiarity with neural networks and machine learning.
- Familiarity with tools such as SQL, R and Python.
- Strong understanding of statistics and linear algebra.
- Strong understanding of hypothesis and model testing, including the ability to identify common model testing errors.
- Experience designing and running A and B tests and drawing insight from them.
- Proficiency in machine learning algorithms.
- Strong analytical skills for sourcing data reliably.

Nice to have, as listed:
- Experience with personalisation or similar machine learning problems.
- Familiarity with big data stacks such as Apache Spark, Hadoop and Redshift.
- Experience working in tech and product teams.

The actual day to day:
- Developing models and running experiments to draw insight from data.
- Improving product usability and identifying growth opportunities.
- Modelling seasonal demand to predict key organisational metrics.
- Designing discount programmes to help resellers sell more.
- Understanding reseller preferences to surface the most relevant products.
- Mentoring junior data scientists on the team.

Location and office reality:
- Bangalore. Office days are not stated in the posting.

Honest fit guidance:
- The mentoring line is mentorship, not people management, so this stays on the individual contributor track.
- The problems are commercial: pricing, discounting, fraud and demand. If you want research, this is not it.
- Business to consumer experience is stated as a preference rather than a requirement, but the whole role is shaped around consumer marketplace behaviour.