Artificial intelligence at ShareChat

We want every Artificial intelligence at ShareChat to be a leader of tomorrow. We want to build an ecosystem where every individual who walks through our doors turns themselves into problem solvers who can lead by example.

The AI Organisation at ShareChat has three streams

Personalisation
Content and User Understanding
Monetization

Personalisation

Focused on improving feed personalization & feed relevance across ShareChat and Moj with the power of AI
In Personalisation, we guide users to discover new trends for ever-evolving interests
Utilize hundreds of ML models to enhance the rankings and fine-tune users every interaction on the platform through the core feed relevance pipeline and a state-of-the-art recommendation engine

Content and User Understanding

Enable automated content moderation to preserve the integrity of the platforms by weeding out different forms of undesirable content (e.g., NSFW, baits, spam).
Make sense of the content posted on ShareChat and Moj in the same way as humans do with the help of multimodal machine learning models, which can process visual, audio, and text data. Automated methods for interpreting semantics and quality of posts to match help us match the content to the right set of users as soon as it gets created.

Monetization

Provide long-term economic value to advertisers and the marketplace runner ShareChat, while creating long term utility for users while at the same time create long term utility for the users
Combine the data on users interaction with ads and the rich preference profiles mined from the users content consumption behavior on the platform to build Ad Tech models, that will help advertisers target their ads effectively to relevant users
The team has developed accurate deep learning models for standard ad use-cases and specialized applications such as CTR and conversion prediction, ad pacing, bid estimation, predicting video ad completion, app installs, lead generation, etc.

Our Leadership

Meet the leaders shaping the journey of ShareChat and Moj

Azitabh Kumar Ajit

Director - Ads Engineering

Yatharth Saraf

Director - AI and Machine Learning, ShareChat

Advancements in AI Research: ShareChat's Comprehensive Paper Collection

Quantifying and Leveraging User Fatigue for Interventions in Recommender Systems

Hitesh Sagtani, Rishabh Mehrotra, Akshat Gupta, Madan Gopal Jhawar

SIGIR'23

CONSEQUENCES -- The 2nd Workshop on Causality, Counterfactuals and Sequential Decision-Making for Recommender Systems

Olivier Jeunen, Thorsten Joachims, Harrie Oosterhuis, Yuta Saito, Flavian Vasile, Yixin Wang

RecSys 2023 (Workshop Proposal)

Scheduling on a budget: Avoiding stale recommendations with timely updates

Robin Verachtert, Olivier Jeunen, Bart Goethals

MLWA, March 2023

3MASSIV: Multilingual, Multimodal and Multi-Aspect dataset of Social Media Short Videos

Vikram Gupta, Trisha Mittal, Puneet Mathur, Vaibhav Mishra, Mayank Maheshwari, Aniket Bera, Debdoot Mukherjee, Dinesh Manocha

CVPR 2022

ADIMA: Abuse Detection In Multilingual Audio

Vikram Gupta, Rini Sharon, Ramit Sawhney, Debdoot Mukherjee

ICASSP 2022

Deep Clustering of Text Representations for Supervision-free Probing of Syntax

Vikram Gupta, Haoyue Shi, Kevin Gimpel, Mrinmaya Sachan

AAAI 2022

Multimodal Video Understanding using Graph Neural Network

Ayush Singh, Vikram Gupta

NeurIPS 2022 (Workshop)

Multilingual Abusive Comment Detection at Scale for Indic Languages

Vikram Gupta, Sumegh Roychowdhury, Mithun Das, Somnath Banerjee, Punyajoy Saha, Binny Mathew, Animesh Mukherjee

NeurIPS 2022 (Dataset Track)

Know your team

Team Spread across India, USA, Canada, Europe & Russia

Team previously made major contributions at Facebook, Linkedln, Spotify, Yahoo, Amazon, Microsoft, Adobe, Oracle, IBM, Uber, Samsung, Visa, to name a few

Top talent from IITs, NITs, BITS, CET, IISC, University of Illinois, Urbana Champaign, ETHZ and Georgia Institute of Technology

Recent publications from the team at top international conferences such as CVPR, ICCV, AAAI, ACL, NAACL, EMNLP, WWW, WSDM, ICWSM, ECIR, ICASSP, INTERSPEECH

University Collaborations with:

UCSD (University of California at San Diego)

UMD (University of Maryland)

UT Austin (University of Texas at Austin)

Team members awarded top honors at ICPC and other prestigious competitive coding competitions

Hear it from our people!

Lead Decision Scientist

Working with ShareChat has been a fantastic experience so far. It is the perfect place for someone looking for a challenging and fast-growing environment. ShareChat is an AI-first org and we are developing many state-of-art solutions in feed personalization, content, and user understanding, monetization, etc. My team is solving interesting problems in RecSys space and Measurement Sciences ranging from the cold start problems of new user feed ranker to more reliable online experiment evaluation and counterfactual evaluation.

- Neeti Pokharna

Staff Engineer

I’ve worked in the IT industry for 5-plus years, focusing on backend and ML/Cloud. My journey in ShareChat has just begun, but I already feel like I’m at the perfect place. Working here includes a remote first approach and excellent work-life balance. The pace of innovation is staggering, and everyone keeps pushing forward each day. With a leadership that strongly believes in and supports AI teams, we can expect even more great things to come!

- Pawel Obrepalski

Staff - Machine Learning Engineer

If your company can build and launch an App in 30 hours that can speed to top the play store chart in just four days of the launch, then you know you’re at the right place! And the array of top-notch ML challenges in developing an effective feed ranking algorithm that works for content across disparate formats like short video, long video, images, and text across I5 Indic languages. Exciting?

- Danish Kalim

Software Development Engineer - 3

I had joined ShareChat as an intern, have been part of multiple pods in the AI team working on entirely different problems and matured to be a better software developer. I have been here with different roles having different expectations yet few things remained constant for example ample opportunities to work on interesting problems at a large scale, need to have ownership, bunch of smartest folks and a really good learning curve. The thrill of making an impact resulting in amazing recognition is one of the best things I love at ShareChat.

Artificial intelligence at ShareChat

The AI Organisation at ShareChat has three streams

Personalisation

Content and User Understanding

Monetization

Our Leadership

Azitabh Kumar Ajit

Yatharth Saraf

Advancements in AI Research: ShareChat's Comprehensive Paper Collection

Quantifying and Leveraging User Fatigue for Interventions in Recommender Systems

Data-Efficient Methods For Improving Hate Speech Detection

Graph Neural Networks for Link Prediction with Subgraph Sketching

RecSys Challenge 2023: Deep Funnel Optimization with a Focus on User Privacy

Practical Bandits: An Industry Perspective

Gradient Boosting Performs Gaussian Process Inference

Which tricks are useful for Learning-to-Rank?

MEMER - Multimodal Encoder for Multi-signal Early-stage Recommendations

A Probabilistic Framework to Learn Auction Mechanisms via Gradient Descent