Description :

As a part of our continued expansion in IT Products and DevOps team, we require the services of a Technical Specialist experienced in building scalable Machine Learning (ML) systems and customizing Deep Learning (DL)/ large language models (LLMs) using automation through MLOps and LLMOps.

Skills Required :

- 9 to 12 Years of experience with proven 7+ years relevant experience in Azure Cloud Deployment, maintaining and managing ML, DL & AI applications.

- Core Skills :

- Cloud : Azure (also familiarity with AWS & GCP)

- Deployment : FastAPI, Docker, Kubernetes, MLOps, MLflow, LLMOps, Azure DevOps CI/CD pipelines, and automation

- Programming : Python, SQL

- ML Libraries : scikit-learn, PyTorch, XGBoost, LightGBM, TensorFlow,

- Deep Learning : Keras API

- GenAI Tools : Hugging Face Transformers, LangChain, OpenAI API

- LLM Models : GPT, LLaMA, Gemini, BERT, Mistral, Falcon

- Fine-Tuning : LoRA, PEFT, RLHF, prompt tuning

- Data Engineering : Pandas, NumPy, Spark

- Knowledge of databases (SQL/NoSQL) and big data technologies (e.g., Hadoop, Spark).

- Knowledge of NLP tasks and evaluation metrics

- Knowledge of Supervised, Unsupervised, Reinforcement learning and computer vision

- Knowledge of Descriptive, Diagnostic, Predictive and Prescriptive Analytics.

- Knowledge of semantic search, RAG pipelines, and vector databases

- Azure ML & Data Engineer certification will be added advantage.

- Strong communication skills

- An analytical mind with problem-solving abilities.

Responsibilities :

As well as possessing the required technical skills, you will be a confident individual capable of working in a busy development environment with a focus on creating secured ML engineering pipelines, to support Data Analyst & Data scientist requirements with required cloud Infrastructure and network architecture.

- Develop AI & ML Models : Design and implement machine learning and deep learning models to solve complex business problems, including but not limited to predictive analytics, natural language processing (NLP), and computer vision.

- Design and implement ML models for classification, regression, recommendation, and forecasting

- Fine-tune transformer-based LLMs (e.g., GPT, LLaMA, Mistral) using LoRA, PEFT, and RLHF techniques

- Develop and optimize prompts for LLMs to improve task-specific performance

- Build end-to-end ML pipelines for data preprocessing, training, evaluation, and deployment

- Data Processing & Analysis : Collaborate with data engineers to gather, clean, and preprocess large datasets, via automated data engineering pipelines (DataOps) for model training and evaluation.

- Monitoring : Monitor deployed models for drift, accuracy, and latency; implement feedback loops and performance benchmarking.

- Model Optimization & Tuning : Continuously improve model performance by fine-tuning hyperparameters, feature engineering, and applying advanced techniques such as transfer learning.

- Deployment : Work on the integration and deployment of ML & AI models into production (Azure, AWS, GCP) environments, ensuring scalability, reliability, and security.

- Collaboration : Work closely with cross-functional teams, including product managers, software engineers, data scientists and other stakeholders, to understand business requirements and translate them into ML & AI solutions.

- Research & Innovation : Stay up to date with the advancements in ML algorithms, latest AI research, DevOps (MLOps, AIOps, LLMOps) practices and trends. Contribute to the development of new algorithms, tools, and techniques that can improve the efficiency and effectiveness of AI models.

- Documentation & Reporting : Document model designs, workflows, and methodologies for internal knowledge sharing and regulatory compliance.

- Work with the leadership to set the standards for Data, ML, LLM engineering practices within the team and support across other disciplines.

- Ensure all release processes, policies and procedures are properly communicated and documented.

- Able to meet tight deadlines, and handle and prioritize simultaneous requests.