All roles / AI
AI Prompt Engineer
Designs and tests the prompts and model workflows a product depends on.
What it pays.
| Experience | Range |
|---|---|
| Fresher, 0 to 1 year | 4.5 to 6 LPA |
| 1 to 3 years | 6 to 10 LPA |
| 3 to 5 years | 9.5 to 14 LPA |
| 5 to 7 years | 14 to 18 LPA |
| 7 to 10 years | 18 to 24 LPA |
Open to freshers. If this is your first job, you are judged on the assessment and not on a CV you have not had a chance to build yet.
Above ten years the range is agreed with the employer case by case. Indicative market ranges for India, and they move with experience, city and employer. Calibre HR does not set salaries, the employer does. The range on this page is not an offer.
What you would do.
- Design prompts and system instructions that hold up
- Build evaluation sets and test against them
- Reduce cost and latency without losing quality
- Document what the model must never do
What you need.
- Hands on work with LLM APIs
- Evaluation method rather than vibes
- Some scripting, and version control
The assessment for this role.
Thirty minutes, one sitting, generated from your own CV. Role knowledge, two judgement scenarios, two questions about work you have listed, and one practical task.
The practical task for this role is usually something like:
Take a prompt that fails on one in ten inputs and say how you would find out why.