Prompt Engineer (AI)

TL;DR

Prompt Engineer (AI): Designing complex system prompts and structured generation flows for AI agents with an accent on Chain-of-Thought techniques, Jinja2 integration, and structured JSON output. Focus on optimizing model performance through A/B testing and analyzing conversation logs to reduce hallucinations and improve agent behavior.

Company

hirify.global is a full-stack development team specializing in large-scale IT products across diverse global industries, including AI-driven solutions.

What you will do

• Design complex structured system prompts using Chain-of-Thought and Few-Shot techniques.
• Engineer prompts to ensure 100% reliable JSON output for API integrations.
• Control agent persona and tone to ensure consistent, empathetic, or professional interactions.
• Implement Jinja2 templates for dynamic data injection into prompts.
• Continuously perform A/B testing to minimize hallucinations and optimize task conversion rates.

Requirements

• Deep expertise with LLMs, including model differences, token limits, and context windows.
• Technical proficiency with JSON, Markdown, and Jinja2 templates.
• Strong analytical mindset for debugging conversation logs and identifying root causes.
• English: B2+ level proficiency required
• Russian: Native or fluent proficiency required

Culture & Benefits

• Remote-first working culture with flexible working hours.
• Opportunity to work on long-term global projects with real-world impact.
• Professional development support including training budgets and paid participation in conferences.
• Access to internal professional communities for knowledge sharing.
• Payment in USD.