Mar 2024 – Jan 2025 · 10 mos · Bengaluru, Karnataka, India · On-site

• Building cutting-edge products in Personalization and Recommendation Systems to transform education.

Apr 2023 – Mar 2024 · 11 mos

• 𝐀𝐝𝐬 𝐏𝐚𝐜𝐢𝐧𝐠
• Worked on the design and integration of PID (Proportional-Integral-Derivative) pacing, a multiplier-based pacing mechanism, aimed at optimizing the budget consumption trend within advertising domain.
• Developed an independent auction simulator replicating the production auction service, encompassing budget allocation, pacing, and auction layers. Leveraged it to run multiple PID experiment simulations, exploring diverse parameter combinations to identify the optimal set.
• The auction simulator also served dual purpose by functioning as a delivery estimator for adops and hypothesis tester supporting new releases across team before an online A/B test - significantly reducing online test time.
• 𝐋𝐨𝐨𝐤𝐚𝐥𝐢𝐤𝐞 𝐌𝐨𝐝𝐞𝐥𝐥𝐢𝐧𝐠
• Led the development and launch of the Lookalike Audience Targeting feature which efficiently expands advertiser audiences by identifying and targeting users with similar traits to an existing base.
• Collaborated with cross-functional teams to bring it live along with audience segments targeting feature on our advertising platform.
• Established the entire modeling paradigm from scratch, leveraging small model(LR), quick inference methodologies to build a robust solution. This involved creating model-feature updation pipelines, audience refresh pipelines, and implementing an incremental model update setup on top of new feedback events.

Aug 2021 – Apr 2023 · 1 yr 8 mos

• Working as a part of Ads & Monetization team built and maintained data pipelines and Looker dashboards from scratch to track key model performances (CTR, CVR, DF) and KPIs used across the entire org to track monetization health.
• Engineered age and gender prediction models, leveraging deep and wide neural networks, with a keen emphasis on clean and modular class-based code. Orchestrated the end-to-end project lifecycle, including the setup for continuous model updates through incremental training. Achieved precise predictions for 60% of MAU in ShareChat. Implemented automated updates via Airflow pipelines for seamless and ongoing model refinement.
• Worked on feature ideation, correlation analysis, feature pipeline development, employed neural network modeling techniques to significantly elevate ShareChat ads CTR model performance, surpassing the control by 5% on registered production CTR.

Jan 2021 – Jul 2021 · 6 mos

• Building interactive dashboards on MSTR for stakeholders tracking key performance indicators.
• Writing complex SQL queries covering different sets of logic and building visualizations around them.
• Working with different tools like Amazon redshift, hive, Apache Airflow etc.
• Key Project –
• 1.) Simulation Engine – The engine simulates key performance indicators like margin upside, GMV upside taking reference from historical data, based on the weights provided by the user to attributes which are indeed responsible for the ranking in which similar product are shown to the user as to what he searched.
• The simulator also provided section where ranking insight for ex. How the position of SKU’s with certain characteristics got affected as a result of weight’s change.
• The simulator also had the built in capacity where the user can compare various iterations which they try out in a session
• Simulator Frontend – HTML, CSS, JavaScript /Backend – python, flask

Jul 2019 – Jul 2019 · 0 mo · Mumbai Area, India

• Worked under the CEO to incorporate machine learning techniques in their existing algorithmic trading system.
• The project was focused on using Machine Learning techniques in order to minimize regular loss observed by the algorithmic system after successive profitable trades.

May 2019 – Jun 2019 · 1 mo · Mumbai Area, India

• Developed and implemented an end to end data science pipeline from scraping to training and predicting the depreciation factor of vehicles.
• Collected data on various margins involved by holding discussions with car dealers and other stakeholders.
• Worked with the CTO in engineering features, focused on results leveraging maximum business value.

Apr 2019 – May 2020 · 1 yr 1 mo

• Working with a team of 15 other executives(board members) for bringing out enthusiasm in the 27,000 student campus for taking up entrepreneurship.

Nov 2018 – Mar 2019 · 4 mos

• We are a student club based in VIT, Vellore which aims at fostering entrepreneurial spirit among young aspirants.
• Guided as a mentor to young fostering minds who brought up ideas for startups , helping them to form a current business model.
• Organised events such as Entrepreneurship Awareness Camp, Startup Pitch & Udan.
• Organised Futurepreneurs a business hack-a-thon consisting current world business problems , formed by extensive research on the current economy and business world
• Organised the E-Summit '19 which encouraged a footfall of around 1500 students along with 27,000 campus students.
• Coordinating the 60 core committee members for bringing up entrepreneur in them & making them work for events to develop best managerial skills.

Jul 2018 – Nov 2018 · 4 mos

• Working in the Research Department. The main task of our department is to work with startup teams, guide them on their road for becoming a successful startup company.