Invent Artificial is an IT services and staffing company. This role sits on a client engagement: an AI interview-preparation platform that runs simulated interviews and gives candidates real-time feedback. You will build and run the LLM systems behind it.
What you will do
- Build and ship GenAI features on top of LLMs such as GPT, Claude, Llama and Gemini.
- Wire models into production services through APIs and SDKs, with LangChain or LlamaIndex where they help.
- Design, test and tune prompts for accuracy, relevance and response quality.
- Build retrieval (RAG) pipelines on vector stores such as Pinecone, FAISS, Chroma or Weaviate.
- Write evaluation harnesses that track accuracy, latency, cost and safety for every model change.
- Cut token spend and latency with caching, batching and model selection.
- Apply guardrails, moderation and access controls, and keep candidate data private.
- Deploy on AWS with containers, CI/CD, versioning and rollback.
- Document the architecture, prompts and APIs so the next engineer can pick them up.
What you need
- A degree in computer science, AI, data science or a related field.
- 2+ years of software engineering, including at least 1 year on LLM or GenAI systems.
- Strong Python and experience building production APIs (FastAPI preferred).
- Hands-on RAG and vector database experience.
- Docker, Kubernetes and Git-based workflows.
- Clear written communication.
Nice to have
- HR-tech, ed-tech or career-platform experience.
- Transformer and NLP fundamentals.
- Fine-tuning, reinforcement learning or multi-agent systems.
Pay and terms
$46 to $50 an hour, W-2, paid hourly and eligible for overtime. This is our good-faith estimate of pay at hire. Full time and fully remote within the United States.
How to apply
Use the form on this page. Attach your resume and a link to something you have built with an LLM: a repo, a demo or a short write-up.