AI Engineer - Intern

NThread is building an AI-native talent intelligence platform designed to help companies make smarter hiring decisions.
We are creating systems that combine AI reasoning, market intelligence and real-world recruiting context to help talent teams understand a role better, navigate the market and identify the right talent faster.
We are an early-stage team, so this is not an internship where you will work on side projects.
You will work directly on the intelligence behind the product.
What You’ll Work On
You will work closely with the founders and engineering team to - our AI intelligence layer - smarter, more reliable and more useful.
Your work will include:
Improve how platform understands roles — job titles, responsibilities, seniority, skills, industries, business models and hiring context.
Build and test AI reasoning frameworks — help define how platform should think through different recruiting and talent-market problems.
Create evaluation datasets — build high-quality examples across roles and functions to test whether answers, recommendations and rankings actually make sense.
Evaluate AI outputs — review responses, search strategies and candidate recommendations and identify where the system is getting things wrong.
Build rulebooks and few-shot examples for different roles, industries and hiring situations.
Experiment with prompts, retrieval and context to improve answer quality and consistency.
Help platform distinguish between a keyword match and genuine contextual fit.
Work with company and talent data to identify patterns that can improve search and recommendation quality.
Create test cases for difficult hiring problems and continuously challenge the system with edge cases.
Work closely with engineering to convert successful experiments into repeatable product behaviour.
We’d Love to Meet You If You
Are deeply curious about LLMs, AI agents and reasoning systems.
Have experimented with APIs, prompts, RAG, embeddings or LLM-based applications.
Can think logically and break ambiguous problems into structured steps.
Enjoy testing systems and figuring out why an AI answer is wrong.
Are comfortable working with datasets using Python, SQL or spreadsheets.
Can research unfamiliar industries, companies and roles quickly.
Have strong attention to detail.
Are comfortable in an early-stage environment where problems are often undefined.
You’ll Be Especially Valuable If You Have
Built something using OpenAI, Anthropic, Gemini or other LLM APIs.
Worked with RAG, vector databases or search systems.
Built evaluation datasets or LLM evaluation frameworks.
Experience with Python and basic data analysis.
An interest in recruitment, talent intelligence, marketplaces or recommendation systems.
If you enjoy asking:
Why did the AI make this decision - and how can we make it better?
you will probably enjoy this role.

Work arrangement
No