Company Research Report: Twelve Labs
Company Overview
- Name: Twelve Labs
- Mission: Build the best video understanding platform driven by leading perceptual-reasoning research.
- Founded By: No specific founding details provided.
- Key People:
- Jae Lee (CEO, Co-founder)
- Aiden Lee (CTO, Co-founder)
- SJ Kim (Head of Engineering, Co-founder)
- Dave Chung (Head of KR Ops, Co-founder)
- Soyoung Lee (Head of Business Development, Co-founder)
- Anthony Giuliani, Esq. (Head of Operations)
- Headquarters:
- San Francisco Office: 55 Green Street, San Francisco, CA 94111
- Seoul Office: 39-11, Itaewon-ro 27-gil, Yongsan-gu, Seoul, Republic of Korea
- Number of Employees: No specific number provided.
- Revenue: No information is available.
- Known For: Video understanding technology, multimodal AI models, semantic video search.
Products
Products Offered
Twelve Labs offers a range of products focused on intelligent video applications powered by their AI technology:
1. APIs for Video Understanding:
- Search: Enables finding any scene in vast video libraries using natural language.
- Generate: Generates text descriptions, summaries, chapters, or hashtags for videos.
- Embed: Provides embeddings for video analysis and applications.
High-Level Description and Key Features
Search
- Purpose: Pinpoint exact moments in videos using natural language.
- Key Features:
- Multimodal search (speech, text, audio, visuals)
- Supports 100+ languages
- Fast and precise indexing and retrieval
Generate
- Purpose: Generate text about video content.
- Key Features:
- Creates various text forms like summaries and highlights
- Promotes interactive and engaging video content understanding
Embed
- Purpose: Generate visual embeddings for video content.
- Key Features:
- Supports multimodal embeddings for video, text, image, and audio
- High performance and reliable
- Integrates with major vector databases
Recent Developments
Recent Developments in the Company
- New Products and Features:
- Marengo 2.6: A state-of-the-art video foundation model for any-to-any search tasks.
- Pegasus-1 Open Beta: A video-language model that enhances video content analysis with 17 billion parameters.
- Embed API Open Beta: Enabling customers to generate multimodal embeddings.
Partnerships
- Partnerships:
- Collaborations with MongoDB, Pinecone, Databricks Mosaic AI, Milvus, LanceDB, ApertureDB for vector storage integrations.
- Major investment from NVIDIA for Series A funding.
Other Developments
- Achievements:
- Recognized as one of the top AI startups by CB Insights.
- Participated in global AI competitions and achieved significant accolades.
-No information on new partnerships or new feature additions beyond those already specified.
Additional Information
Security and Compliance
- Security Framework:
- Adheres to defense in-depth security principle
- Applies multiple layers of security controls
- SOC2 compliant and working towards ISO 27001 compliance
Recognition and Awards
- Recognitions:
- Named to Fast Company's Most Innovative Companies of 2023
- Industry recognition for advancements in multimodal AI models
Investment and Financial News
- Series A Funding: Raised $50 million with participation led by NVIDIA and NEA.
This report captures a comprehensive snapshot of Twelve Labs using the available data, highlighting its leadership in video understanding technology and its extensive lineup of AI-driven video products and services. The focus remains aligned with the company's mission of advancing multimodal AI applications in video content analysis and retrieval.
