Vectara is on a mission to provide a scalable platform for neural information retrieval, helping our customers build advanced language understanding into the next generation of software systems. Our stellar team includes CEO Amr Awadallah, cofounder of Cloudera, experts in neural IR, distributed database systems, and search platforms from organizations including Google, Google Research, Elasticsearch, Databricks, and Twilio.
This position is with Vectara’s machine learning team. You’ll be a good fit if:
1. You have a passion for working on language understanding with deep neural methods.
2. You have prior industry expertise, or have just completed graduate studies in this area.
3. Excited about working in a collaborative team setting.
4. You are open to living and working in Silicon Valley. While we support working from home, you should be open to the idea of working as a team in our office in Palo Alto, California three days a week.
Come help us build the future of information retrieval!
Ideally, you will have expertise in one or more of the following areas:
- Experience with multimodal and multilingual implementation.
- Design of dual-encoder models, or transformer based rerankers, or LLMs for information retrieval.
- Experience with a major deep learning toolkit like Pytorch or JAX.
- Design and development of online learning systems.
- Neural approaches to Open domain QA and Zero-shot IR.
- Familiarity with sparse and dense IR methods.
- Machine learning techniques, including, but not limited to, deep learning.
The following are plusses, but not required:
- Published research.
- Vector database expertise, including performance measurement and quantization techniques.
What We Do
Vectara is an LLM-powered Answer as a Service. The platform provides a #chatgpt like service which enables you to have a conversation with your own data. Our 100% neural natural language understanding (NLU) technology provides radical increases in semantic relevance. Instead of a list of results ala legacy keyword-based systems, we provide a direct answer to the user's input prompt. Vectara’s complete neural pipeline is available as a service through an API-first platform that allows developers to easily add semantic answer serving within their applications.