PeopleofSuperintelligence

Builder

Alykhan Tejani

Alykhan Tejani is an AI/ML engineer whose work spans computer vision, deep-learning software, and recommender systems. His papers cover 3D hand-pose estimation, GAN-based image super-resolution, and the PyTorch library; he also helped start PyTorch-Ignite. In 2020, he co-authored studies on online advertising and privacy-aware recommendations for Twitter's Home Timeline. In 2025, he co-authored ShareChat's post on low-cost GPU training for recommendation models. As of October 2026, he is a senior staff machine-learning engineer at ShareChat.

Now at
ShareChat
Works
20

Publication numbers from OpenAlex, as of 15 October 2026. Citations and h-index are hidden because this OpenAlex record may be incomplete or mix in other authors.

Timeline

  1. 2014

    Estimating 3D hand pose

    In 2014, he co-authors a CVPR paper that presents a latent regression forest for estimating 3D articulated hand posture from depth images.

  2. 2017

    Improving image resolution

    He co-authors a 2017 CVPR paper that uses a generative adversarial network for single-image super-resolution. Its method aims to restore finer texture in enlarged images.

  3. 2017–2018

    Starting PyTorch-Ignite

    In 2017, he starts PyTorch-Ignite with Francisco Massa and Joost van Amersfoort. The project publishes its first public release in June 2018.

  4. 2019

    Co-authoring PyTorch

    In 2019, he co-authors the PyTorch paper, which describes its imperative Python style, debugging approach, runtime implementation, and architecture.

  5. 2020

    Recommendation research

    In 2020, he co-authors a study that frames display advertising as a contextual bandit and tests exploration methods with offline simulations and online production traffic. A separate 2020 paper describes a privacy-aware challenge for Twitter's Home Timeline and introduces a dataset of 160 million Tweets with engagement data.

  6. January 2025

    Training ShareChat recommenders

    In January 2025, he co-authors a ShareChat engineering post on training recommendation models with low-cost GPUs. The post describes large models for personalized recommendations.

  7. As of October 2026

    Senior staff engineer at ShareChat

    As of October 2026, he is a senior staff machine-learning engineer at ShareChat. PyTorch-Ignite is a high-level library for training and evaluating neural networks in PyTorch.

Sources and credits

Photo: We have no published photo of Alykhan Tejani yet. Send us a free-licensed photo.

  1. Latent Regression Forest: Structured Estimation of 3D Articulated Hand Posture openaccess.thecvf.com. Checked 15 October 2026.
  2. Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network openaccess.thecvf.com. Checked 15 October 2026.
  3. About us — PyTorch-Ignite v0.4.3 Documentation docs.pytorch.org. Checked 15 October 2026.
  4. PyTorch: An Imperative Style, High-Performance Deep Learning Library proceedings.neurips.cc. Checked 15 October 2026.
  5. Deep Bayesian Bandits: Exploring in Online Personalized Recommendations arxiv.org. Checked 15 October 2026.
  6. Privacy-Aware Recommender Systems Challenge on Twitter's Home Timeline arxiv.org. Checked 15 October 2026.
  7. Scaling Recommender System Training on Low-Cost GPUs at ShareChat sharechat.com. Checked 15 October 2026.
  8. alykhantejani (Alykhan Tejani) · GitHub github.com. Checked 15 October 2026.