Head of Artificial Intelligence (AI Head)

Location:

Employment Type:

Reporting To:

Department:

defining, building, and implementing AI-driven solutions across the enterprise

AI strategy, use-case identification, model development, deployment, and governance

retail, supply chain, manufacturing, sales, marketing, finance, and customer service

Key Responsibilities & Accountabilities

• AI Strategy & Roadmap
• Define and execute the enterprise AI strategy and roadmap aligned with business objectives and digital transformation goals.
• Identify high-impact AI/ML use cases across functions (forecasting, optimization, automation, personalization, fraud detection, predictive maintenance, etc.).
• Prioritize initiatives based on business value, feasibility, and ROI.
• AI Solution Development & Deployment
• Lead end-to-end AI solution lifecycle: problem definition, data preparation, feature engineering, model development, testing, deployment, and monitoring.
• Design and implement ML, Deep Learning, NLP, Computer Vision, and Generative AI solutions using industry-standard frameworks and libraries (e.g., Python, TensorFlow, PyTorch, scikit-learn, XGBoost, LightGBM).
• Hands-on experience with NLP toolkits and transformer models (Hugging Face Transformers, spaCy, BERT, GPT-family, LLaMA), embeddings, prompt engineering, and retrieval-augmented generation (RAG) for LLM-based solutions.
• Computer vision experience with OpenCV, Detectron2, EfficientDet, YOLO and CNN/transformer-based architectures for image/video tasks.
• Ensure seamless integration of AI models with ERP, CRM, WMS, OMS, e-commerce platforms, data warehouses and expose ML capabilities via REST/gRPC APIs or microservices.
• Champion production readiness: containerization (Docker), orchestration (Kubernetes), API design, load testing, latency optimization, and edge/ONNX/TensorRT deployments where applicable.
• Data, MLOps & Platforms
• Define AI architecture including data pipelines, model training environments, cloud platforms, feature stores, and MLOps frameworks. Provide clear patterns for training, validation and production serving.
• Data engineering and pipeline examples: Apache Spark (batch), Kafka / Debezium (streaming), Flink (stream processing), Airflow / Argo Workflows (orchestration), and ETL/ELT with tools like dbt.
• Feature store and data quality: Feast, Tecton; data validation and testing with Great Expectations and data monitors.
• MLOps platforms & practices (explicit examples): experiment tracking (Weights & Biases, MLflow, Neptune), model registries (MLflow Registry, DVC), CI/CD for ML (Jenkins, GitHub Actions, GitLab CI, Tekton), workflow orchestration (Kubeflow Pipelines, Argo, Prefect), and reproducible pipelines (Pachyderm).
• Model serving & inference: KServe, Seldon Core, BentoML, TorchServe, TensorFlow Serving; low-latency inference optimizations using ONNX, TensorRT, NVIDIA Triton, and edge deployments where needed.
• Distributed training & acceleration: PyTorch DDP, Horovod, DeepSpeed, TensorFlow MultiWorker; GPU/TPU provisioning and management on cloud or on-prem clusters.
• Cloud AI platforms & infra automation: AWS (SageMaker, EKS, ECS, S3, RDS), GCP (Vertex AI, GKE, BigQuery), Azure (Azure ML, AKS); infrastructure as code using Terraform, CloudFormation, Helm charts.
• Observability, monitoring & model governance: Prometheus, Grafana, ELK stack, Evidently, WhyLabs, Fiddler for model monitoring, drift detection, data quality alerts, and automated retraining triggers.
• Vector databases & retrieval stores for embeddings: Pinecone, Milvus, Weaviate, Qdrant; typical use with RAG pipelines and semantic search architectures.
• Security, compliance & cost optimization: secrets management (HashiCorp Vault, cloud KMS), network policies, IAM best practices, and cost monitoring/rightsizing for GPUs & storage.
• Typical production ML workflow (example pattern): data ingestion (Kafka) -> ETL & feature engineering (Spark + dbt) -> experiments (PyTorch/TensorFlow + W&B) -> model registry (MLflow) -> CI/CD (GitHub Actions + Argo) -> serving (Seldon/BentoML on Kubernetes) -> monitoring (Prometheus + Evidently) -> retraining pipeline (Airflow/Kubeflow).
• Business Enablement & Adoption
• Partner with business leaders to translate business problems into AI-driven solutions.
• Drive AI adoption and change management, ensuring users trust and effectively use AI outputs.
• Enable AI-powered dashboards, decision-support tools, and automation systems.
• Governance, Ethics & Compliance
• Establish AI governance frameworks, including data privacy, explainability, bias mitigation, and ethical AI principles.
• Ensure compliance with data protection laws, healthcare/pharma regulations, and ISO standards.
• Work closely with Information Security teams to ensure secure and compliant AI implementations.
• Team Leadership & Capability Building
• Build and lead high-performing AI, Data Science, and ML Engineering teams (onshore/offshore).
• Define skill frameworks, hiring plans, training programs, and succession planning.
• Foster a culture of innovation, experimentation, and continuous learning.
• Vendor & Partner Management
• Evaluate and manage AI vendors, platforms, cloud providers, and system integrators.
• Negotiate SLAs, performance metrics, and commercial terms.
• Ensure optimal cost-to-value outcomes for AI investments.
  

• Number of AI use cases deployed to production
• Business value and ROI delivered through AI initiatives
• Model accuracy, stability, and adoption rates
• Time-to-deploy AI solutions
• Compliance, security, and audit outcomes
• AI platform scalability and cost optimization
  

Qualifications

• Bachelor’s degree in Computer Science, Engineering, Mathematics, or a related field.
• 10+ years of overall professional experience, including a minimum of 6 years in AI/ML leadership roles managing cross-functional teams and delivering enterprise solutions.
• Proven track record of designing and implementing enterprise-scale AI/ML solutions that delivered measurable business impact (examples: forecasting, personalization, automation, predictive maintenance, fraud detection).
• Strong hands-on experience with Python and SQL and working knowledge of ML/DL frameworks such as TensorFlow or PyTorch and scikit-learn.
• Operational experience deploying models to production using containerization and orchestration (Docker, Kubernetes) and exposing models via APIs/microservices.
• Familiarity with at least one major cloud provider and its ML services (AWS/GCP/Azure) and practical experience with model training and inference on cloud infrastructure.
• Demonstrated expertise in MLOps practices (CI/CD for ML, model versioning/registry, experiment tracking, monitoring) and the ability to implement robust production pipelines.
• Strong software engineering discipline: version control (Git), testing, code review practices, and API design.
• Proven ability to translate business problems into AI solutions and to communicate with senior stakeholders and cross-functional teams.
  

Preferred Qualifications (Optional / Nice-to-have)

• Master’s degree or PhD in AI, Data Science, Machine Learning, or Analytics.
• Experience with big data and data engineering technologies (Apache Spark, Kafka, Flink), cloud data warehouses (Snowflake, Redshift, BigQuery) and orchestration tools (Airflow).
• Hands-on familiarity with MLOps/platform tooling: MLflow, DVC, Kubeflow, Argo, Seldon, WhyLabs; infrastructure automation with Terraform and Helm.
• Experience with advanced ML topics: distributed training (GPU/TPU), ONNX/TensorRT optimizations, latency and cost optimization techniques.
• Practical experience with NLP/LLMs (Hugging Face, transformer models), embeddings, RAG architectures, vector databases (Milvus, Pinecone, Weaviate) and prompt engineering for production use cases.
• Experience in computer vision solutions (OpenCV, Detectron2, YOLO, ViT) where applicable.
• Knowledge of model interpretability and governance tools/methods (SHAP, LIME), bias detection & mitigation approaches, and privacy-preserving techniques (differential privacy, anonymization).
• Domain experience in healthcare/pharma, retail, manufacturing, BFSI, or supply chain is a plus.
• Track record of publications, patents, or speaking at conferences is advantageous.
  

• [This section highlights concrete technical keywords and grouped tooling for search and candidate matching — examples provided to improve discoverability and clarify expected stacks]
• Languages & Core: Python (primary), SQL, Bash; secondary: R, Scala, Java
• ML / DL Frameworks: PyTorch, TensorFlow, JAX, scikit-learn, XGBoost, LightGBM, CatBoost
• NLP & LLMs: Hugging Face Transformers, spaCy, BERT, GPT-family, LLaMA, SentenceTransformers, tokenizers, prompt engineering, RAG patterns
• Computer Vision: OpenCV, Detectron2, YOLO family, EfficientDet, Vision Transformers (ViT)
• Data & Big Data: Apache Spark, Kafka, Flink, Hive, Presto, dbt; cloud warehouses: Snowflake, Redshift, BigQuery
• Feature Stores & Data Quality: Feast, Tecton, Great Expectations
• MLOps & Experimentation: MLflow, Weights & Biases, Neptune, DVC, Guild; experiment tracking and model registry tools
• Orchestration & Pipelines: Airflow, Kubeflow Pipelines, Argo Workflows, Prefect
• Model Serving & Inference Platforms: Seldon Core, KServe, BentoML, TorchServe, TensorFlow Serving, NVIDIA Triton
• Vector & Embedding Stores: Pinecone, Milvus, Weaviate, Qdrant
• Containerization & Orchestration: Docker, Kubernetes (EKS/GKE/AKS), Helm
• Infra Automation & CI/CD: Terraform, CloudFormation, Jenkins, GitHub Actions, GitLab CI, Tekton
• Distributed Training & Acceleration: PyTorch DDP, Horovod, DeepSpeed; GPU/TPU management
• Optimization & Model Formats: ONNX, TensorRT, quantization, pruning
• Monitoring & Observability: Prometheus, Grafana, ELK, Evidently, WhyLabs, Fiddler
• Datastores & Caching: PostgreSQL, MySQL, MongoDB, Redis, Elasticsearch
• Security & Governance: HashiCorp Vault, cloud KMS, IAM best practices, privacy-preserving tooling
• Testing & Quality: pytest, tox, Great Expectations for data, integration & end-to-end ML tests
• Example typical production ML stack (concise): Python + PyTorch/TensorFlow, feature store (Feast), experiment tracking (W&B/MLflow), model registry (MLflow), CI/CD (GitHub Actions + Argo), serving (Seldon/BentoML on Kubernetes), observability (Prometheus + Grafana + Evidently), vector DB (Pinecone) for RAG scenarios.
  

• Strong business-first mindset with ability to translate AI into measurable outcomes
• Strategic thinking with hands-on technical depth
• Excellent stakeholder management and executive communication
• High ethical standards and governance orientation
• Ability to operate in fast-paced, transformation-led environments