Experience
- At least 5 years
- The applicants should have experience in the following business area(s): IT Enabled Service
Education
- Bachelor of Science (BSc) in Mathematics, Computer Science & Engineering
Requirements
- At least 5 years
- The applicants should have experience in the following business area(s): IT Enabled Service
Additional Requirements
Hands-on machine learning engineering OR equivalent experience (including self-directed research, open-source model work, competition results or industry work without a research title).
Direct experience training or adapting language or foundation models at any scale OR equivalent experience.
Strong implementation in Python with PyTorch, JAX or equivalent, with understanding of what happens beneath the framework
Rigorous evaluation and ablation design
Diagnosing training pathologies and regressions
Building defensible training data pipelines
Distributed training: mixed precision, sharding, checkpointing and multi-node debugging
Agentic post-training, tool use or reinforcement learning from feedback
Inference optimisation, quantisation, distillation and serving efficiency
Publications, open-source models or datasets, or contributions to widely used ML libraries
Responsibilities & Context
Run model training programs: continued pretraining of base models, supervised fine-tuning, preference optimization and post-training for domain and agentic workloads.
Improve model quality through data, objectives, architecture, optimization and training recipes, with clear attribution of which lever produced the change.
Design evaluations and ablations that distinguish real improvement from noise, including evaluation of agentic behavior and long-horizon tasks.
Diagnose and fix training pathologies: instability, divergence, numerical issues, data contamination, catastrophic forgetting and capability regressions.
Build training data pipelines: filtering, deduplication, mixture design and synthetic generation — with defensible provenance.
Work with systems engineers to scale training efficiently across in-house hardware, and with product engineers to bring models into deployment.
Establish reusable practice for experimentation and model release.
Job Other Benifits
- Provident fund,Medical allowance,Mobile bill,Weekly 2 holidays,Insurance,Gratuity
- Lunch Facilities: Partially Subsidize
- Salary Review: Yearly
- Festival Bonus: 2