Career & Salary Map — CapabilityNext

Read this first. Salary figures are indicative ranges for India, compiled from public salary guides (September 2026) — they vary a lot by city, company type and the individual. Canada ranges will be added and shown by location later. “Entry” assumes you finish with a portfolio and capstone, not a zero-experience application. Course fees shown as ₹[XX,XXX] are placeholders — replace them with your real prices. Time-to-ready assumes part-time study (8–12 h/week). Nothing here is a guarantee of a job or a salary; it is a planning aid. Sources are listed at the bottom.

The mapping

What you spend → what you earn → how long

Read each block top to bottom: the skills a job needs, the one CapabilityNext course per skill, its length and price — then the salary that whole path targets. India figures; switch the experience band above the table.

Experience band

Scroll the table sideways →

Salary	Job profile	Skill required	Our course (one per skill)	Duration	Price
₹6–12 LPAper year	AI / Machine Learning Engineer	Python, data tooling & the ML maths	Python & Maths for Machine Learning	8 weeks	₹[fee]
		Supervised & unsupervised ML, evaluation	Introduction to Machine Learning	8 weeks	₹[fee]
		Deep learning with PyTorch	Deep Learning Foundations with PyTorch	9 weeks	₹[fee]
		Full path	—	~25 weeks~6–7 months part-time	₹[total]
₹8–14 LPAper year	LLM / Generative-AI Engineer	Python foundations (if new to it)	Python & Maths for Machine Learning	8 weeks	₹[fee]
		Prompting, RAG, agents & guardrails	LLM Application Engineering	10 weeks	₹[fee]
		Retrieval & vector search in production	RAG & Vector Search in Production	6 weeks	₹[fee]
		Fine-tuning (LoRA) & model evaluation	Fine-tuning & Model Evaluation	7 weeks	₹[fee]
		Full path	—	~31 weeks~7 months part-time	₹[total]
₹6–10 LPAper year	MLOps Engineer	Data pipelines, orchestration & feature stores	Data Engineering for Machine Learning	9 weeks	₹[fee]
		Deploy, monitor & retrain models (CI/CD)	MLOps on the Cloud	9 weeks	₹[fee]
		LLM serving & inference at scale	LLMOps & Inference at Scale	6 weeks	₹[fee]
		Full path	—	~24 weeks~5–6 months part-time	₹[total]
₹5–10 LPAper year	Data Engineer	SQL, Python & data modelling	Python & Maths for Machine Learning	8 weeks	₹[fee]
		Pipelines, orchestration & data quality	Data Engineering for Machine Learning	9 weeks	₹[fee]
		Cloud deployment & ML-platform basics	MLOps on the Cloud	9 weeks	₹[fee]
		Full path	—	~26 weeks~6 months part-time	₹[total]
₹7–13 LPAper year	Computer Vision / NLP Engineer	Deep-learning foundations	Deep Learning Foundations with PyTorch	9 weeks	₹[fee]
		Vision models — detection, segmentation	Applied Computer Vision	9 weeks	₹[fee]
		Language models — transformers, NER, seq2seq	NLP with Transformers	8 weeks	₹[fee]
		Multimodal & vision-language systems	Multimodal & Vision-Language Systems	6 weeks	₹[fee]
		Full path	—	~32 weeks~7–8 months part-time	₹[total]
₹5–9 LPAper year	DevOps → Platform / MLOps	ML data pipelines & feature stores	Data Engineering for Machine Learning	9 weeks	₹[fee]
		Model serving, monitoring & retraining	MLOps on the Cloud	9 weeks	₹[fee]
		Full path	—	~18 weeks~4 months part-time	₹[total]

₹[fee] / ₹[total] are placeholders — drop in your real course prices. Salary bands are indicative India ranges from public guides (see sources below); “entry” assumes a finished portfolio and capstone. Generative-AI roles carry a ~25–40% premium over generalist ML; Bengaluru, Hyderabad, Pune and Gurugram sit at the top of every band.

AI / Machine Learning Engineer

Builds, trains, evaluates and ships models. The broad, portable role most people target first.

₹6–12 LPAIndia, entry (with portfolio)

₹15–30 LPAIndia, 2–5 years

Skills employers screen for

Python, NumPy / pandas, the ML maths
Supervised & unsupervised learning, evaluation
Deep learning with PyTorch
Turning a model into a service

Courses that cover them

Python & Maths for ML · Foundations · 8 wk · ₹[fee]
Introduction to Machine Learning · Foundations · 8 wk · ₹[fee]
Deep Learning Foundations with PyTorch · Foundations · 9 wk · ₹[fee]

Path: ~25 weeks + the 2-week foundations sprint ≈ 6–7 months part-time · total ₹[total] · target entry ₹6–12 LPA.

LLM / Generative-AI Engineer

Designs, builds and ships LLM products — RAG, agents, evaluation, guardrails. Currently the highest-paid entry point.

₹8–14 LPAIndia, entry

₹18–35 LPAIndia, 2–5 years

Skills employers screen for

Prompt design & structured output
Retrieval-augmented generation, vector search
Tool use, agents, evaluation suites
Fine-tuning (LoRA), guardrails, cost / latency

Courses that cover them

LLM Application Engineering · GenAI & LLM · 10 wk · ₹[fee]
RAG & Vector Search in Production · GenAI & LLM · 6 wk · ₹[fee]
Fine-tuning & Model Evaluation · GenAI & LLM · 7 wk · ₹[fee]
Python & Maths for ML · Foundations · 8 wk · ₹[fee] (if you're new to Python)

Path: ~23 weeks (or ~31 with foundations) ≈ 5.5–7.5 months part-time · total ₹[total] · target entry ₹8–14 LPA.

MLOps Engineer

Packages, deploys, monitors and retrains models. Where software / DevOps engineers most naturally cross into AI.

₹6–10 LPAIndia, entry

₹12–22 LPAIndia, 2–5 years

Skills employers screen for

CI/CD for ML, model registries
Pipelines & orchestration, feature stores
Containerised serving, autoscaling
Monitoring, drift detection, retraining

Courses that cover them

Data Engineering for Machine Learning · Data & MLOps · 9 wk · ₹[fee]
MLOps on the Cloud · Data & MLOps · 9 wk · ₹[fee]
LLMOps & Inference at Scale · Data & MLOps · 6 wk · ₹[fee] (for LLM-serving roles)

Path: ~18 weeks (add ~6 for LLMOps) ≈ 5–7 months part-time · total ₹[total] · target entry ₹6–10 LPA.

Data Engineer

Builds the pipelines and stores that everything else depends on. The lowest-friction entry into the field.

₹5–10 LPAIndia, entry

₹10–20 LPAIndia, 2–5 years

Skills employers screen for

SQL & Python, data modelling
Batch & streaming pipelines, orchestration
Data quality, testing, cloud warehouses
Feature stores for ML

Courses that cover them

Python & Maths for ML · Foundations · 8 wk · ₹[fee]
Data Engineering for Machine Learning · Data & MLOps · 9 wk · ₹[fee]
MLOps on the Cloud · Data & MLOps · 9 wk · ₹[fee] (to move toward ML-platform roles)

Path: ~17 weeks (26 with MLOps) ≈ 4–6 months part-time · total ₹[total] · target entry ₹5–10 LPA.

Computer Vision / NLP Engineer

A deep-learning specialisation — image models or language models — on top of the ML-engineer base.

₹7–13 LPAIndia, entry

₹16–30 LPAIndia, 2–5 years

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

Courses that cover them

Deep Learning Foundations with PyTorch · Foundations · 9 wk · ₹[fee]
Applied Computer Vision or NLP with Transformers · CV & NLP · 8–9 wk · ₹[fee]
Multimodal & Vision-Language Systems · CV & NLP · 6 wk · ₹[fee] (advanced)

Path: ~26–32 weeks ≈ 6–8 months part-time (Python & Maths first if needed) · total ₹[total] · target entry ₹7–13 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.

₹5–9 LPAIndia DevOps, entry

₹12–22 LPAIndia MLOps, mid

What you already have

CI/CD, containers, Kubernetes, IaC
Cloud, observability, on-call

What you add

ML pipelines, feature stores, model registries
Model serving, batching, drift & retraining

Courses that cover the gap

Data Engineering for Machine Learning · 9 wk · ₹[fee]
MLOps on the Cloud · 9 wk · ₹[fee]

You can usually skip Python & Maths for ML — confirm on your advisor call.

Path: ~18 weeks ≈ 4–5 months part-time · total ₹[total] · MLOps roles typically pay ₹3–6 LPA more than the equivalent DevOps role.

The maths

Spend vs. earn vs. time

A worked example, once you drop in real fees. Say an LLM-Engineer path costs ₹X and takes 6 months part-time. If it moves you into a role at ₹10 LPA that you would not otherwise have reached, the course cost is recovered in roughly X ÷ (10,00,000 ÷ 12) months of that salary — often a single-digit number of months — and everything after that is upside.

What the number does not capture: whether you'd have got there anyway, how long the job search takes, and that individual offers vary widely. Treat this page as a way to compare paths, not as a promise about your outcome.

Method & sources

Ranges are the rough consensus across the public salary guides below, read in September 2026, weighted toward the entry and 2–5-year bands. We did not scrape LinkedIn or job boards — their terms prohibit it, and aggregated guides give steadier numbers. Refresh this page every quarter.

India

Glassdoor India — MLOps Engineer, Data Engineer
Futurense — MLOps, ML Engineer, Data Engineer, AI Engineer, LLM Engineer
Instahyre — DevOps Engineer, AI/ML Engineer
Recrew — LLM Engineer, Data Engineer, Platform & DevOps
upGrad — MLOps salary, Data Engineer · KnowledgeHut — MLOps Engineer · taggd — AI Engineer

Canada — not shown yet; sources kept for when location-based figures are added

Indeed Canada — ML Engineer, DevOps Engineer
Glassdoor Canada — ML Engineer, DevOps Engineer, Senior DevOps
PayScale — Data Engineer (CA) · levels.fyi — ML Engineer (CA) · Morgan McKinley — ML Engineer Toronto
Robert Half — DevOps Engineer Toronto · Coursera — DevOps salary · Jobicy — AI Engineer (CA)

Provider names identify the source, not an endorsement. Figures on those pages change; always check the current version before relying on it.

What each AI role pays — and what it costs to get there

What you spend → what you earn → how long

AI / Machine Learning Engineer

LLM / Generative-AI Engineer

MLOps Engineer

Data Engineer

Computer Vision / NLP Engineer

DevOps Engineer → Platform / MLOps

Spend vs. earn vs. time

Method & sources

India

Canada — not shown yet; sources kept for when location-based figures are added

Map your path in 20 minutes