Meesho
About this role
This is a senior data science seat on Meesho's advertising platform, and the problems are the classic
hard ones in ad systems: real time bidding, ad ranking, budget pacing and working out whether any of
it actually worked. You design the bidding and pacing algorithms that decide how ad spend is
allocated, build the experimental frameworks that measure ad effectiveness and return, and translate
what you find into product and engineering changes. The methods named are auction theory, predictive
modelling, causal inference and reinforcement learning, built in PyTorch or TensorFlow with Python
and big data tooling. It is a role with business consequence attached: ads revenue is a direct line
on the P and L. Apply if you have shipped a model that moved money rather than one that scored well
offline.
hard ones in ad systems: real time bidding, ad ranking, budget pacing and working out whether any of
it actually worked. You design the bidding and pacing algorithms that decide how ad spend is
allocated, build the experimental frameworks that measure ad effectiveness and return, and translate
what you find into product and engineering changes. The methods named are auction theory, predictive
modelling, causal inference and reinforcement learning, built in PyTorch or TensorFlow with Python
and big data tooling. It is a role with business consequence attached: ads revenue is a direct line
on the P and L. Apply if you have shipped a model that moved money rather than one that scored well
offline.
Who this is for
Requirements. A B.Tech in Computer Science, Machine Learning or a related field, with at least 6+
years of experience in AI/ML research. Experience in auction theory, predictive modelling, causal
inference and reinforcement learning. Expertise in PyTorch or TensorFlow, and proficiency in Python
as well as big data technologies. A history of mentoring junior data scientists and comfort driving
multiple problem statements at once.
What you would do. Design and develop bidding, budget pacing and related algorithms that optimise ad
spend and return. Create frameworks and experimental designs to measure ad effectiveness and ROI.
Work with product and engineering to turn business requirements into ML solutions. Communicate the
long term roadmap, insights and recommendations to technical and business leadership. Mentor junior
data scientists and provide technical guidance. Lead research initiatives in machine learning,
auction theory and advertising technology.
Team context. Meesho's data science group works across fraud detection, inventory optimisation and
platform vernacularisation, so the AdTech team sits inside a broader group solving commerce problems
for a mostly non metro user base.
Location and terms. Bangalore, marked on-site, Full Time Employee, on the Data Science team.
Who this is for. A data scientist with six or more years whose work has been in ranking, auctions,
bidding or another domain where a model's output is money. The auction theory and causal inference
requirements are specific: general ML experience without exposure to either will be a stretch. The
mentoring line is stated as a requirement rather than a responsibility, so bring examples. Note the
title says Principal but the stated bar is 6 years, which is lower than most Principal titles on this
board and is why it sits in the 6 year group rather than the senior one.
years of experience in AI/ML research. Experience in auction theory, predictive modelling, causal
inference and reinforcement learning. Expertise in PyTorch or TensorFlow, and proficiency in Python
as well as big data technologies. A history of mentoring junior data scientists and comfort driving
multiple problem statements at once.
What you would do. Design and develop bidding, budget pacing and related algorithms that optimise ad
spend and return. Create frameworks and experimental designs to measure ad effectiveness and ROI.
Work with product and engineering to turn business requirements into ML solutions. Communicate the
long term roadmap, insights and recommendations to technical and business leadership. Mentor junior
data scientists and provide technical guidance. Lead research initiatives in machine learning,
auction theory and advertising technology.
Team context. Meesho's data science group works across fraud detection, inventory optimisation and
platform vernacularisation, so the AdTech team sits inside a broader group solving commerce problems
for a mostly non metro user base.
Location and terms. Bangalore, marked on-site, Full Time Employee, on the Data Science team.
Who this is for. A data scientist with six or more years whose work has been in ranking, auctions,
bidding or another domain where a model's output is money. The auction theory and causal inference
requirements are specific: general ML experience without exposure to either will be a stretch. The
mentoring line is stated as a requirement rather than a responsibility, so bring examples. Note the
title says Principal but the stated bar is 6 years, which is lower than most Principal titles on this
board and is why it sits in the 6 year group rather than the senior one.
Apply on company site
Opens jobs.lever.co, the employer's own application page. Applying is always free.