Reply.io is a fast-growing SaaS company helping businesses automate and scale their sales outreach.

We're looking for a Prompt Engineer to work with large language models (GPT, Claude, and others) — enhancing their performance through prompt tuning and conducting research using AI coding agents.

This is hands-on, focused prompt work — sometimes repetitive, always precise — with direct impact on how Jason AI and our other AI-powered features perform.

If your success formula is: 👉 Right model + Right prompt structure + Right evaluation + Continuous iteration — you'll thrive here.

🎯 What you'll do:

• Refine existing prompts — continuously improve quality across our AI features, from simple categorisation to complex, highly personalised email sequence generation, including addressing minor bugs in the results
• Develop new prompts for upcoming AI tools and workflows aimed at making customers' work easier and more efficient
• Use AI coding agents to research, prototype, and validate prompt approaches
• Test prompts on newly released models (Claude, GPT, and others) — compare performance, document trade-offs, recommend upgrades
• Partner with sales, growth, and dev teams to make sure AI outputs match real business needs
• Embrace the detail-oriented, sometimes repetitive nature of prompt tuning — consistent quality matters more than novelty here
• Stay ahead of LLM advancements — new models, new techniques, new evaluation methods

🧩 Requirements:

• 0.5–1 year of experience as a Prompt Engineer, understanding of AI agent workflows, or light scripting/development background (will be a plus — candidates without prior experience are also welcome to apply)
• Passion for AI and hands-on experience with LLMs (GPT, Claude, or similar)
• Prompt tuning expertise — you've optimised prompts and can articulate what worked, what didn't, and why
• Attention to detail and patience for focused, methodical iteration — the role involves repetitive refinement with consistent quality
• Strong communication — you can translate between business needs and LLM behaviour, working across sales, growth, and dev teams
• English proficiency (B2+) — written and spoken
• Willingness to learn — you actively follow AI trends and test on new models as they release

🎁 What we offer:

• Ownership of prompts that ship into customer-facing AI features used by sales teams every day
• Work with cutting-edge LLMs — Claude, GPT, and whatever ships next; premium API access covered
• High autonomy — no micromanagement, no bureaucracy, real ownership
• 100% remote with minimal meetings
• Coverage of professional courses, gym memberships, or therapy sessions
• AI-first culture — we actively use and cover costs of premium tools
• 15 paid vacation days, 🇺🇦 national holidays & ~10 days of Christmas vacation
• Sick leaves without limitations
• Access to internal training, literature, and knowledge sessions

If you love prompting, think in iterations, and are excited to shape how Jason AI performs — we'd love to hear from you!