PARTNERS / INTERNSHIPS
Internships
Learn through a concrete contribution to AI-native work.
At inAi, internships are considered individually rather than through one fixed programme. We consider proposals from students and people seeking a structured early-career placement when their interests and abilities connect with a meaningful area of work, the project has clear learning value, and we have enough capacity to support it properly.
Possible directions include AI-native products, software for agents, Open Source, research and evaluation, public AI explanation, and product design. These are areas in which a placement may be shaped, not a permanent list of open roles.
What makes a useful internship
An internship should be serious without treating the intern as a finished professional. It should give the candidate practical experience, context, and feedback while producing a bounded contribution to work that actually matters.
A coherent objective
The placement should be organised around one clear question, project, or deliverable. The scope may be modest, but it should be concrete enough to finish and useful enough to matter.
Learning through real work
The intern should develop skills and judgment by working through an actual problem, not by observing from the side or receiving a collection of unrelated tasks. The work should stretch the candidate without asking them to operate as an unsupervised employee.
Context and feedback
We only confirm a placement when we can provide enough context and feedback for the work to be educational. The exact rhythm depends on the project, but the intern should understand what they are trying to achieve, where they can make decisions, and where review is needed.
Where an internship could fit
These are possible directions, not permanent openings. What is viable at a particular time depends on the work in progress, the candidate’s abilities and learning goals, the scope of the proposed project, and our capacity to support it properly.
Products, design, and engineering
Contribute to a scoped product problem: a prototype, interface, evaluation, integration, developer tool, workflow study, or bounded technical improvement. The point is to make AI useful inside a real product or process, not to add AI for its own sake.
Software for agents and Open Source
Explore software that AI agents can understand and use, or improve a public project through code, testing, documentation, examples, or developer experience. The exact project would depend on the current maturity of the work and whether it can support a meaningful placement.
Research and evaluation
Work on a clearly defined question through literature review, experiments, evaluations, research notes, or structured analysis connected to AGI as a System, agentic decision systems, knowledge creation, or AI in business operations.
AI for Everybody and technical communication
Research and explain an important AI topic for broad readers through writing, visuals, interactive material, or source-backed guidance. The goal is clarity without simplifying the subject until it becomes misleading.
Business · Consumers · Products for Agents · Open Source · AGI / Research · AI for Everybody
What inAi can offer
Depending on the project, an internship can place you close to real questions in AI-native products, agent-facing software, Open Source tools, research, or public AI explanation. A confirmed placement should give you a defined area of work, the context needed to understand it, and feedback on the decisions and output that matter.
Where the project allows it, the result may become a public contribution or a reusable artifact. Where it does not, the placement should still leave you with a coherent piece of work, a clearer understanding of the problem, and specific feedback on how you approached it.
What we look for
You do not need to arrive as a finished professional. We look for specific interest in one part of inAi, evidence that you can make, investigate, explain, or improve something, honesty about your current level, clear communication, and the ability to carry a bounded piece of work through.
A university project, personal prototype, repository, research note, design case, article, experiment, or another piece of work can be enough if you can explain your contribution, the decisions you made, what did not work, and what you learned. A small piece of work you understand deeply is more useful than a polished application filled with generic claims.
Example project directions
These examples show the scale and shape of a useful internship project. They are not advertised openings.
Product workflow prototype
Study one product or operational workflow, identify where AI could improve it, and produce a small prototype, evaluation plan, or design proposal with documented assumptions and limits.
Agent-ready documentation study
Examine how software documentation changes when the reader may be an AI agent as well as a human. A possible outcome could be a short research note, a documentation pattern, and recommendations for an agent-facing tool.
Open Source developer experience
Improve the documentation, examples, tests, or contribution path of one public project. A possible outcome could be a focused pull request or a documented improvement that another builder can use.
AI for Everybody explainer
Research one public question about AI and turn it into an accessible, source-backed guide, visual explanation, or interactive element for non-specialist readers.
Propose an internship
We do not need a long cover letter. A short, specific message is more useful.
Please include:
who you are, what you study or are training in, and your current level;
your approximate dates, expected duration, location, and any requirements from your school, university, or programme;
the part of inAi that interests you and why;
what you can already do and what you want to learn;
one or two examples of relevant work, with your own contribution explained;
an optional project direction or problem you would like to explore.
Do not include confidential, personal, customer, supplier, candidate, security-sensitive, or otherwise restricted material in an initial message. Links to work you are allowed to share are sufficient.
For universities, schools, and programmes
If you are contacting inAi on behalf of an institution, include the programme name, expected format and duration, formal agreement or reporting requirements, relevant deadlines, and the person responsible for coordinating the placement.
Example first message
Subject: Internship — [Area] — [Name]
Hello, my name is [name]. I am studying [programme or field] at [institution], and I am looking for an internship around [dates and expected duration].
I am particularly interested in [inAi area] because [specific reason]. I have worked on [example], where my contribution was [brief explanation]. I would like to develop [skill or area of judgment], and I think I could contribute through [project idea or capability].
My programme requires [agreement, reporting, or other requirement], if applicable. My work is available at [links].
I would be interested in discussing whether there is a fit.
How a placement is defined
Internships are considered individually rather than through one fixed format. We first look for a meaningful fit between the candidate, an actual area of work, and our capacity to support the placement. Where that fit exists, the next conversation defines the project, learning goals, timing, duration, working arrangement, academic documentation, and other terms before either side commits.
A suitable project or enough support capacity may not be available when a proposal arrives. We would rather say that clearly than create a placement without meaningful work.
Start with a specific proposal
Send your introduction to careers@inai.world with the subject line Internship — [Area] — [Name].
Looking for a different relationship?
For employment, contract work, or broader collaboration, visit Work with us. For formal research collaboration, visit Academia / Research. For a bounded public contribution, visit Contributions.
Work with us · Academia / Research · Contributions · How We Build