AIApply is hiring a remote Prompt Engineer to own the quality of every model-generated word our users see. Our products write resumes, cover letters, interview answers and outreach for job seekers worldwide — and the prompt layer is the product.
You will design, test and version prompts across multiple model providers, build evaluation sets that catch regressions before users do, and turn fuzzy quality complaints into measurable improvements. This is a rare role where linguistic craft, product judgement and engineering rigour meet.
Fully remote, async-friendly, with fast shipping cycles and direct impact on retention and conversion.
Design, iterate and version production prompts for resume generation, cover letters, interview coaching and job matching.
Build golden datasets and automated evaluations (rubric scoring, LLM-as-judge, pairwise comparisons) that quantify output quality across models.
Run structured A/B tests on prompt variants and report on quality, latency and cost trade-offs.
Tune structured outputs, JSON schemas and function-calling contracts so downstream code never breaks on malformed responses.
Benchmark and route between models — GPT, Claude, Gemini and open-weight alternatives — choosing the right model for each task.
Reduce hallucination, tone drift and formatting errors through better context construction, few-shot design and guardrails.
Work with engineering to move winning prompts into production safely, and with support to close the loop on real user complaints.
2+ years working directly with LLMs in a product, research or engineering context, including shipped prompt work in production.
Exceptional written English and a strong editorial eye — you can tell good copy from adequate copy and explain why.
Comfortable with light scripting (Python or TypeScript) to build eval harnesses, batch runs and analysis notebooks.
Deep familiarity with prompting techniques: few-shot, chain-of-thought, structured outputs, context compression, retrieval grounding and guardrails.
Analytical mindset — you instrument, measure and iterate rather than relying on intuition.
Nice to have: experience in recruiting, careers or HR tech; fine-tuning experience; multilingual prompt work.
Must-haves: 2+ years hands-on LLM prompt work in production, outstanding written English, basic Python or TypeScript scripting, and experience building or running systematic model evaluations. Portfolio or examples of prompt work strongly preferred.
-
Prompt engineering
-
Large Language Models (LLMs)
-
LLM evaluation and benchmarking
-
Python
-
Structured outputs and JSON schema
-
Few-shot and chain-of-thought prompting
-
A/B testing and experimentation
-
OpenAI, Anthropic and Gemini APIs
-
Technical writing and editing
-
Fully remote — work from anywhere
-
Competitive salary plus equity
-
Flexible, async-first schedule
-
Annual learning and AI tooling budget
-
Home office stipend
-
Generous paid time off
-
High autonomy and visible product impact
