Prompt Engineer

Creuto; Khordha; Sur place; INR 600 000 - 1 000 000; Plein temps

Résumé du poste

Creuto is seeking a Prompt Engineer in India to design, test, and refine prompts powering AI features across our products. You will collaborate with AI engineers, product managers, and designers to ensure prompts are reliable, safe, and scalable.

The role emphasizes building prompt libraries, evaluating prompt quality, and running systematic experiments to push performance while addressing failure modes and hallucinations.

Qualifications

• 1 to 3 years of experience working with LLMs such as GPT, Claude, Gemini, or open-source models.
• Think analytically about language and can reason precisely about why a prompt produces a given output.
• Are highly systematic and treat prompt engineering as an empirical discipline, not creative guesswork.
• Have strong written communication skills and an intuition for how models interpret instructions.
• Are comfortable working in Python to build prompt testing pipelines and evaluation scripts.

Responsabilités

• Design, iterate, and optimise prompts for LLM-powered features across product surfaces.
• Build and maintain a prompt library with versioning, documentation, and benchmarks.
• Develop evaluation frameworks to measure prompt quality, reliability, and output consistency.
• Run experiments to test prompt variations and identify high-performing approaches.
• Collaborate with product and engineering teams to integrate prompts into production systems.
• Identify failure modes and edge cases in LLM outputs and engineer solutions.
• Stay current on prompting techniques including chain-of-thought, RAG, few-shot, and agentic patterns.
• Collaborate on fine-tuning decisions and model selection for use cases.
• Document prompting strategies and share learnings across the team.

Connaissances

LLMs

Prompt testing

Python

Outils

LangChain

LlamaIndex

Description du poste

Prompt Engineer

DEPARTMENT

Artificial Intelligence

EMPLOYMENT TYPE

Full Time

LOCATION

India

EXPERIENCE

1 + Years

About the Role

We are looking for a Prompt Engineer with 1 to 3 years of experience to design, test, and refine prompts that power AI features across our products. You will work closely with AI engineers, product managers, and designers to craft prompts that are reliable, safe, and deliver consistently great outputs to users at scale.

In This Role, You Will

• Design, iterate, and optimise prompts for LLM-powered features across our product surfaces.
• Build and maintain a prompt library with versioning, documentation, and performance benchmarks.
• Develop evaluation frameworks to measure prompt quality, reliability, and output consistency.
• Run systematic experiments to test prompt variations and identify the highest-performing approaches.
• Work with product and engineering teams to integrate prompts into production systems reliably.
• Identify failure modes, hallucination patterns, and edge cases in LLM outputs and engineer solutions.
• Stay current on prompting techniques including chain-of-thought, RAG, few-shot, and agentic patterns.
• Collaborate with AI engineers on fine-tuning decisions and model selection for specific use cases.
• Document prompting strategies and share learnings across the team.

You Might Thrive in This Role If You

• Have 1 to 3 years of experience working with LLMs such as GPT, Claude, Gemini, or open-source models.
• Think analytically about language and can reason precisely about why a prompt produces a given output.
• Are highly systematic and treat prompt engineering as an empirical discipline, not creative guesswork.
• Have strong written communication skills and an intuition for how models interpret instructions.
• Are comfortable working in Python to build prompt testing pipelines and evaluation scripts.

Bonus If You Have

• Experience with LangChain, LlamaIndex, or similar LLM orchestration frameworks.
• Familiarity with vector databases and RAG architectures.
• Knowledge of model fine-tuning and when prompting alone is insufficient.
• Experience red-teaming LLMs for safety and reliability.