Krishna Chaitanya

Lead ML-driven optimization across Nexverse’s core ad exchange, focusing on floor price algorithms, dynamic margin optimization, and intelligent request / traffic shaping to improve yield and profitability.​

Design and productionize real-time models that set impression-level price floors and margins, balancing net revenue, win rate, fill rate, and advertiser performance across diverse inventory a

Led the design and deployment of ML-driven margin and pricing optimization models for InMobi’s ad exchange, delivering millisecond-level latency while operating at internet scale on Azure Databricks and PySpark.​

Drove annual margin improvement of approximately 5.5 million USD and revenue uplift of around 14 million USD through margin optimization models, directly improving profitability and unit

• Conceptualized and designed data science solutions to enhance ad delivery efficiency at Pokkt.
• Developed and deployed five machine learning models for predicting click-through rates, addressing extreme class imbalance.
• Designed a machine learning model for optimizing bid call decisions, ensuring millisecond-level latency and improved response rates.

• Collaborated with Arima Research to tackle complex data science challenges, enhancing operational efficiency.
• Developed and implemented a resume matching algorithm, significantly improving candidate-job alignment.

• Led a Data Science team at Airpush, focusing on the implementation of ML algorithms for projects like Fraud Detection and Bid Optimization.
• Developed a machine learning model that optimized bidding for cost-per-install campaigns, saving $500,000 annually.
• Created a deep learning solution for brand safety, effectively detecting inappropriate content in child-targeted apps.

• Led the design and implementation of a user profiling system for over a billion users, extracting demographic insights from mobile app installations.
• Utilized advanced technologies including Apache Spark, Stanford NLP, and Python to analyze user data effectively.
• Delivered actionable insights on demographics and interests, enhancing targeted marketing strategies for clients.

• Developed and deployed three machine learning models for predicting click-through rates (CTR) in mobile ad campaigns.
• Addressed extreme class imbalance in datasets, improving model accuracy despite low CTR (~5%).
• Ensured millisecond-latency inference using technologies like Python, Keras, and Apache Spark.
• Contributed to Airpush, a leading mobile advertising platform, enhancing user

• Developed predictive models for sales forecasting of store items for Aditya Birla Group in Bengaluru.
• Utilized advanced time series techniques to analyze historical sales data, enhancing accuracy in predictions.
• Collaborated with cross-functional teams to implement data-driven strategies, improving inventory management.

• Collaborated with cross-functional teams to develop a robust front-end user interface and back-end server APIs.
• Enhanced the framework for financial and insurance products, improving efficiency and user experience.
• Utilized technologies such as GWT, HTML, and CSS to deliver high-quality software solutions.

Studied courses not only in physics, mathematics and chemistry but also from varied fields like psychology, economics, sociology, management, sanskrit language etc...

Integrated M.Sc. in Physics with coursework spanning physics, mathematics, chemistry, and varied fields like psychology, economics, sociology, management, and Sanskrit.