AI is not a side project at VirtualStation. Teams use it to turn meetings into summaries and action items, run standups without another call, brainstorm ideas, and see what is blocking work. These features run on real conversations, every day, for real teams.
We are looking for an early-career engineer who wants to build with large language models in production. You will work on features end to end, from prompt and data design to APIs, evaluation and monitoring, alongside our product and engineering team.
What you will do
- Build and improve AI features such as meeting summaries, action items, AI standups and brainstorming.
- Design prompts, structured outputs and retrieval pipelines, and make them reliable on messy real-world input such as transcripts.
- Set up evaluations so we can measure quality, compare models and catch regressions before release.
- Write the backend services and APIs that connect models to the product, with attention to cost, speed and privacy.
- Keep up with new models and techniques, test them quickly, and recommend what is worth adopting.
- Work with QA and product to understand where AI output falls short, and fix it.
What you bring
- 0 to 2 years of experience, including internships or substantial personal projects. Freshers with strong projects are welcome.
- Solid programming skills in Python or JavaScript/TypeScript.
- Hands-on experience calling LLM APIs and building something real with them.
- Understanding of prompt engineering, embeddings and retrieval augmented generation (RAG).
- Comfort with REST APIs, JSON, Git and basic databases.
- A careful, testing mindset. You check whether the output is actually right.
Nice to have
- Experience with speech to text, transcription or summarisation.
- Experience with agent frameworks, tool calling, or vector databases.
- A GitHub profile, demo or write-up of AI projects you have built.