AI / Machine Learning Engineer
Builds, trains, evaluates and ships models. The broad, portable role most people target first.
Skills employers screen for- Supervised & unsupervised learning, evaluation
- Deep learning with PyTorch
- Turning a model into a service
- Python, NumPy / pandas, the ML maths
- Linear algebra & calculus for ML
- Probability & statistics for ML
- Supervised & unsupervised ML, evaluation
- Bias, leakage & explainability
Path: ~25 weeks + the 2-week foundations sprint ≈ 6–7 months part-time · total ₹1,06,000 + GST · target entry ₹6–12 LPA.
Computer Vision / NLP Engineer
A deep-learning specialisation — image models or language models — on top of the ML-engineer base.
Skills employers screen for- Deep learning, CNNs and transformers
- Detection / segmentation, or sequence tasks
- Data pipelines, augmentation, transfer learning
- Evaluation and deployment of the model
- Vision models — detection, segmentation
- Language models — transformers, NER, seq2seq
- Multimodal & vision-language systems
- Tokenisation & text embeddings
- Image–text embeddings & retrieval
- Document understanding & OCR pipelines
Courses that cover them- Deep Learning Foundations with PyTorch · Foundations · 9 wk
Covers: Deep learning, CNNs and transformers - Applied Computer Vision · CV & NLP · 9 wk
Covers: Detection / segmentation, or sequence tasks, Data pipelines, augmentation, transfer learning, Vision models — detection, segmentation - NLP with Transformers · CV & NLP · 8 wk
Covers: Evaluation and deployment of the model, Language models — transformers, NER, seq2seq, Tokenisation & text embeddings - Multimodal & Vision-Language Systems · CV & NLP · 6 wk
Covers: Multimodal & vision-language systems, Image–text embeddings & retrieval, Document understanding & OCR pipelines
Path: ~26–32 weeks ≈ 6–8 months part-time (Python & Maths first if needed) · total ₹1,24,500 + GST · target entry ₹7–13 LPA.
Data Engineer
Builds the pipelines and stores that everything else depends on. The lowest-friction entry into the field.
Skills employers screen for- SQL & Python, data modelling
- Batch & streaming pipelines, orchestration
- Data quality, testing, cloud warehouses
- Feature stores for ML
- Pipelines, orchestration & data quality
- Python, NumPy / pandas, the ML maths
Courses that cover them- Data Engineering for Machine Learning · Data & MLOps · 9 wk
Covers: SQL & Python, data modelling, Batch & streaming pipelines, orchestration, Data quality, testing, cloud warehouses, Feature stores for ML, Pipelines, orchestration & data quality - Python & Maths for Machine Learning · Foundations · 8 wk
Covers: Python, NumPy / pandas, the ML maths
Path: ~17 weeks (26 with MLOps) ≈ 4–6 months part-time · total ₹81,000 + GST · target entry ₹5–10 LPA.
DevOps Engineer → Platform / MLOps
If you already do DevOps, you keep most of your toolkit and add the ML-specific parts. The fastest cross-over path.
Skills employers screen for- CI/CD, containers, Kubernetes, IaC
- Cloud, observability, on-call
- ML pipelines, feature stores, model registries
- Model serving, batching, drift & retraining
- Cloud deployment & ML-platform basics
- Deploy, monitor & retrain models (CI/CD)
Courses that cover them- MLOps on the Cloud · Data & MLOps · 9 wk
Covers: CI/CD, containers, Kubernetes, IaC, Cloud, observability, on-call, ML pipelines, feature stores, model registries, Cloud deployment & ML-platform basics, Deploy, monitor & retrain models (CI/CD) - LLMOps & Inference at Scale · Data & MLOps · 6 wk
Covers: Model serving, batching, drift & retraining
Path: ~18 weeks ≈ 4–5 months part-time · total ₹53,000 + GST · MLOps roles typically pay ₹3–6 LPA more than the equivalent DevOps role.
LLM / Generative-AI Engineer
Designs, builds and ships LLM products — RAG, agents, evaluation, guardrails. Currently the highest-paid entry point.
Skills employers screen for- Prompt design & structured output
- Retrieval-augmented generation, vector search
- Tool use, agents, evaluation suites
- Fine-tuning (LoRA), guardrails, cost / latency
- Tool use & agentic workflows
- Prompting, RAG, agents & guardrails
- Retrieval & vector search in production
- Fine-tuning (LoRA) & model evaluation
- Retrieval evaluation & grounded citations
- LLM serving & inference at scale
- Quantisation, distillation & inference optimisation
Courses that cover them- LLM Application Engineering · GenAI & LLM · 10 wk
Covers: Prompt design & structured output, Tool use & agentic workflows, Prompting, RAG, agents & guardrails - RAG & Vector Search in Production · GenAI & LLM · 6 wk
Covers: Retrieval-augmented generation, vector search, Retrieval & vector search in production, Retrieval evaluation & grounded citations - Fine-tuning & Model Evaluation · GenAI & LLM · 7 wk
Covers: Tool use, agents, evaluation suites, Fine-tuning (LoRA), guardrails, cost / latency, Fine-tuning (LoRA) & model evaluation - LLMOps & Inference at Scale · Data & MLOps · 6 wk
Covers: LLM serving & inference at scale, Quantisation, distillation & inference optimisation
Path: ~23 weeks (or ~31 with foundations) ≈ 5.5–7.5 months part-time · total ₹1,15,000 + GST · target entry ₹8–14 LPA.
MLOps Engineer
Packages, deploys, monitors and retrains models. Where software / DevOps engineers most naturally cross into AI.
Skills employers screen for- CI/CD for ML, model registries
- Pipelines & orchestration, feature stores
- Containerised serving, autoscaling
- Monitoring, drift detection, retraining
- Deploy, monitor & retrain models (CI/CD)
- Cloud deployment & ML-platform basics
- Model serving, monitoring & retraining
- LLM serving & inference at scale
- Quantisation, distillation & inference optimisation
Courses that cover them- MLOps on the Cloud · Data & MLOps · 9 wk
Covers: CI/CD for ML, model registries, Containerised serving, autoscaling, Monitoring, drift detection, retraining, Deploy, monitor & retrain models (CI/CD), Cloud deployment & ML-platform basics, Model serving, monitoring & retraining - Data Engineering for Machine Learning · Data & MLOps · 9 wk
Covers: Pipelines & orchestration, feature stores - LLMOps & Inference at Scale · Data & MLOps · 6 wk
Covers: LLM serving & inference at scale, Quantisation, distillation & inference optimisation
Path: ~18 weeks (add ~6 for LLMOps) ≈ 5–7 months part-time · total ₹99,500 + GST · target entry ₹6–10 LPA.
Salary figures are indicative, compiled from published salary guides and refreshed quarterly. See each course page for its syllabus and delivery options.