Genie Ubers Gen Ai On Call Copilot
- Source
- uber.com
- Author
- Uber Engineering 1
- Date
Context
Genie lets engineers seek answers from internal documentation before waiting for an on-call colleague. Uber Engineering built the assistant to retrieve relevant passages from its knowledge sources, give those passages to a and request answers with source links. Users can ask follow-up questions or escalate to human support. Uber's account describes this retrieval-based workflow.
Uber treats documentation quality as part of the system, because retrieval can only be as good as the documents it finds. A separate evaluator scores source documents and suggests improvements. Users can also mark an answer as fully resolving a question, only partly helpful, not helpful, or not something Genie can assist with, and Uber records those ratings alongside the conversation.
- LLM — A large language model — the neural network behind tools like Claude and ChatGPT, trained on huge amounts of text to predict what comes next.
- RAG — Retrieval-augmented generation — fetching relevant documents first and pasting them into the model's context so it answers from your data instead of memory.
- fine-tuning — Taking a trained model and training it a bit more on your own examples so it gets better at one specific job.
- hallucination — When a model states something false with full confidence — inventing facts, citations, or APIs that don't exist.
Checking sign-in…
Loading comments…