Description :
The role involves working on next-generation AI solutions using modern LLM architectures, deep learning frameworks, and large-scale data pipelines. The selected candidates will engage in building advanced models, optimizing training and inference performance, and deploying models into production environments that handle real-world workloads.Professionals in this role will collaborate with cross-functional teams, contribute to architectural decisions, mentor junior engineers, and ensure that ML/LLM projects are delivered with high quality, reliability, and scalability.
Key Responsibilities
- Design, develop, and optimize Large Language Models (LLMs) and other deep learning architectures such as Transformers, encoder-decoder models, and attention-based systems.
- Build scalable training pipelines using frameworks such as PyTorch, TensorFlow, Keras, Scikit-Learn, DeepSpeed, or Hugging Face.
- Work on fine-tuning, pre-training, and domain adaptation of LLMs for various use cases including text generation, summarization, classification, sentiment analysis, and RAG-based retrieval systems.
- Develop robust and efficient ML model training and inference workflows, ensuring low latency, high throughput, and cost-efficient operation.
- Collaborate with data engineering and DevOps teams to deploy LLM and ML pipelines into cloud-based or hybrid production environments.
- Manage, lead, or guide technical teams responsible for delivering ML/LLM models, ensuring milestone achievement and project alignment with business goals.
- Build scalable APIs, model-serving systems, and inference endpoints for high-performance production use cases.
- Conduct research on the latest advancements in LLMs, distributed training, parameter-efficient techniques, quantization, and model optimization.
- Implement best practices in model versioning, experiment tracking, and model lifecycle management.
- Work with large datasets, create preprocessing pipelines, and perform efficient feature extraction and embedding generation.
- Ensure production-grade reliability, monitoring, observability, and continuous improvement of deployed models.
Required Skills : Learning & Deep Learning :
- Strong foundation in ML algorithms, deep learning architectures, NLP techniques, and model evaluation methodologies.
- Hands-on experience with modern architectures including Transformers, Encoder-Decoder models, and LLMs.
Frameworks & Tools
Proficiency with at least two of the following :
- PyTorch
- TensorFlow
- Keras
- Hugging Face Transformers
- Scikit-Learn
- DeepSpeed
Programming & Engineering
- Strong expertise in Python for ML model development, automation, and pipeline creation.
- Experience working with distributed training, GPU-based model development, and optimization techniques.
Production & Deployment
- Experience deploying ML/LLM models in production environments using cloud-native tools or scalable architecture patterns.
- Understanding of CI/CD processes, API development, model serving frameworks, and monitoring tools.
Leadership & Collaboration
- Ability to manage or mentor teams, provide technical guidance, and oversee end-to-end ML delivery cycles.
- Strong problem-solving, analytical thinking, and cross-team collaboration skills.
Preferred Experience (Good To Have)
- Experience working with RAG systems, vector databases, and embeddings.
- Exposure to MLOps platforms and workflow systems.
- Experience with optimization libraries, quantization techniques, and multi-GPU training setups.
- Research or publication experience in LLM, NLP, or deep learning fields.
Who Should Apply
Candidates who can work independently, bring clarity to complex problem statements, and take ownership of delivering high-performance ML/LLM solutions in a distributed team environment. Individuals who continuously stay updated with the latest LLM research and enjoy working on challenging, large-scale AI systems will excel in this role.(ref:hirist.tech)
