Vectara

We founded Vectara in 2022 to build an AI Assistant/Agent solution for enterprises. We provide a serverless, end-to-end Retrieval-Augmented-Generation (RAG) platform. This platform combines multi-lingual hybrid (semantic and lexical) information retrieval with AI-generated responses and actions. Developers can optimize its behavior through APIs instead of modifying core algorithms. Enterprise customers and technology partners embed our GenAI platform into their applications using plug-n-play API integrations.

Our focus is on accuracy, explainability, and secure access control. We work to eliminate “hallucinations,” ensure results are explainable, and provide secure access. We also prioritize testability, scalability, reliability, availability, and resilience to prompt attacks, along with copyright protection and bias mitigation, to ensure our platform is trustworthy for enterprise applications.

Deployment options include on-premise, SaaS, or customer-managed VPC. The platform is model-agnostic, supporting Bring Your Own Model (BYOM), and is delivered as a fully hosted service. It uses context-engineering techniques for relevant retrieval and supports multimodal data and complex documents like text, tables, and images. Our policy-led enforcement offers continuous governance, detecting and correcting hallucinations, and enforcing built-in policy and brand controls at runtime. We designed Vectara for enterprise-scale, from pilot to production, to meet security, reliability, and operational readiness requirements.

We specialize in Neural Search, RAG, Natural Language Processing, Natural Language Understanding, Machine Learning, Large Language Models, and Generative AI. We hold SOC 2 Type 2 compliance, HIPAA compliance, GDPR, and AWS FTR certifications. Security features include OAuth & API Keys, Encryption (Self-Managed Keys), and Access Controls (RBAC/ABAC). Our case studies include Anywhere, which used our platform for a centralized RAG command center; Broadcom, which modernized its customer experience with conversational AI; and IEEE, which created a semantic search engine.

Our mission is to help people find meaning in vast amounts of data using accurate generative AI. We were early innovators in Retrieval Augmented Generation (RAG) and continue to improve our retrieval, generation, and hallucination detection models.

Our culture is built on shared values we group under “ONE Focused”:

• One Team: We collaborate openly across individuals and teams, welcome new ideas, and promote accountability.
• Nimbleness: We believe in continuous improvement, learn from ideas that don’t work, and strive to be agile, quality-focused learners and problem-solvers.
• Enabling Success: We listen to customers, partners, and each other, embrace opportunities to solve problems, and support them as customer-focused problem-solvers.

We value diversity, equity, inclusion, and belonging. We aim to attract talented individuals and offer fair compensation based on job level and location, welcoming people from all backgrounds. Our team maintains a positive, inclusive culture that values talent from all backgrounds.

Our team consists of over 50 “Vectarans,” including engineers, scientists, and problem-solvers. As a global team, we value diverse perspectives and collaborate across time zones. Our culture is diverse, collaborative, and inclusive. Our founders began working on AI at Google.

We offer benefits and perks, which vary by region. These include medical, life, and disability insurance, plus an employee assistance program. Employees receive maternity/paternity leave, bereavement leave, and unlimited vacation time. We offer flexible working hours, a monthly phone/internet stipend, and a home office stipend for remote employees. Headquarters employees get free gym access, snacks, and beverages. We also organize an annual global summit and have Minimal Meeting Fridays.