AI engineers with work you can trace from research to implementation

An Indian foundation model company developing language models for local and enterprise use. Earlier experience includes a global consumer electronics research organisation.

• His professional site identifies contributions to 30B and 105B mixture of experts models and work on supervised fine tuning and reinforcement learning.
• His CV describes asynchronous training pipelines that separate generation, reward computation and policy updates.
• Coauthored published research on transformer training stability, providing a separate technical record beyond the job description.

An Indian foundation model business, with an academic research record focused on low resource languages. His work spans translation, speech and multilingual model evaluation.

• Coauthored IndicTrans2, a multilingual machine translation project supporting the 22 scheduled Indian languages.
• His research record includes IndicLLMSuite, connecting language model development with the data resources required for Indian languages.
• Subsequent work covers crosslingual evaluation and translation of long, structured material, extending the scope beyond short benchmark sentences.

A global creative software company with research and product teams in Bengaluru. His specialism is computer vision, including controllable animation and generative visual experiences.

• His professional research page connects CVPR 2022 image animation work to the Moving Elements feature released in a consumer photo editing product in 2023.
• The associated research project exposes the authors and technical approach, making the research contribution independently inspectable.
• His record bridges academic computer vision and a commercial editing workflow, giving the profile a concrete product connection.

An independent ML practitioner with earlier experience at a global developer platform and a global accommodation marketplace. His public work focuses on assessing model behaviour and improving AI applications.

• Coauthored CodeSearchNet, a research benchmark spanning six programming languages and linking natural language queries with source code.
• The accompanying engineering article explains why semantic code search needs evaluation beyond ordinary keyword matching.
• His later public writing covers evaluation datasets, error analysis and model based judging for AI applications.

An open source AI platform supporting model development and sharing. His public professional profile places him in Switzerland and identifies language model post training work.

• Coauthored research introducing a common evaluation framework and hub for machine learning models.
• His public model platform profile exposes an ongoing connection to the organisation's research and engineering ecosystem.
• His published interview discusses language models, reinforcement learning from human feedback and the practical tooling required to work with them.

A Canadian enterprise AI business building language model products. This is a founder and research leadership example, with a technical publication record predating the company.

• Coauthored Dynamic Routing Between Capsules, a neural network research paper with two other researchers.
• The publication identifies his contribution to collaborative research on representing and routing visual features.
• His subsequent company role connects research experience with building an enterprise AI business; that founder scope is distinct from an individual contributor vacancy.

An applied AI research laboratory in Queensland, following leadership at a global data science competition platform.

• Coauthored a 2018 paper on language model fine tuning for text classification, evaluated across six classification tasks.
• The paper reported relative error reductions of 18 to 24 percent on most of those datasets. Those are paper specific results, not a claim about production models.
• Cofounded an applied AI research laboratory focused on turning research into useful end user products, with earlier leadership at a data science competition platform.