Machine Learning Research Engineer, Apple Intelligence (International Languages), at Apple, Bengaluru
Job Description
This NLP research engineer job in Bangalore is with Apple’s Product Engineering International team, part of Software Engineering. You will build generative language models for popular international languages in Apple Intelligence. Experience with Indian languages is a plus.
At a glance
- Position: Machine Learning Research Engineer, Apple Intelligence (role 200614863-0321)
- Where: Bengaluru, India
- Team: Product Engineering International, Software Engineering (Machine Learning and AI)
- Experience: 5–8 years of machine learning expertise
- Posted: 8 September 2026
- Apply: open until filled (re-checked 5 November 2026)
What the machine learning research engineer does
- Build and optimise generative models for a range of popular international languages
- Own the whole pipeline: data preprocessing, model architecture, training, evaluation and deployment
- Make sure the models adapt well to each language and culture
- Work on text generation features such as context-augmented rewriting, safety-controlled composition, free-form text transformation and personalised interactions
- Present plans, progress, results and demos regularly to the team
Why this role matters
Most large language models are trained mainly on English. Making them work well in Hindi, Tamil or Bengali needs real research on data, tokenization and evaluation. This team builds that for key international markets, and the work ships to Apple devices.
What Apple is looking for
- 5–8 years of machine learning expertise
- Strong programming in Python, Objective-C or Swift
- Good exposure to deep learning libraries such as PyTorch and TensorFlow
- Thorough experience with common NLP methods: tokenization, language modelling, text decoding and text classifiers
Nice to have
- Hands-on ML work: writing optimizers, regularisers and loss functions, tuning training, and reproducing and revising state-of-the-art research experiments
- Fixing accuracy and speed bottlenecks in distributed training and inference, and benchmarking model compression methods
- High-quality code in large repositories, and automated training and evaluation pipelines
- Multimodal modelling
- NLP modelling for Indian languages
Pay and location
The posting does not state pay. The role is based in Bengaluru and is not listed as a work-from-home role.
How to apply
Apply through the official Apple job posting. Apple’s posting does not give a closing date, so the role is open until filled; ResearchJobs.in will re-check this listing on 5 November 2026. Check the posting for location and work-authorisation details before you apply.
See all our artificial intelligence jobs, or browse more research jobs on ResearchJobs.in.
Hiring institution: Apple
Official advertisement: jobs.apple.com
How to prepare for this application
- Prepare to explain one research result you reproduced, and what you changed when it did not match the paper
- Revise tokenization for Indic scripts: subword vocabularies, script mixing and transliterated text
- Be ready to discuss decoding methods (beam search, sampling, constrained decoding) and when each fails
- Think about how you would evaluate a rewriting model in a language with few public benchmarks
- Brush up on distributed training and model compression: data and tensor parallelism, quantization and distillation
About Apple
Apple is a US technology company that makes the iPhone, Mac, iPad and other products. Apple Intelligence is its set of generative AI features across its devices. Apple has engineering teams in Bengaluru and Hyderabad.
