AI21 Labs, a renowned NLP company based in Tel Aviv, has unveiled its latest offering, Contextual Answers. This plug-and-play generative AI engine is designed to assist enterprises in extracting value from their data assets. By providing a dedicated API that seamlessly integrates with digital assets, organizations can leverage large language model (LLM) technology on their selected data to deliver a conversational experience. This eliminates the need for multiple teams or software systems, enabling business employees and customers to obtain the desired information effortlessly.
Unlike previous solutions, Contextual Answers is a ready-to-use technology that does not require extensive effort or resources. AI21 Labs has developed a comprehensive plug-and-play capability, optimizing each component to ensure that clients achieve industry-leading results without relying on AI, NLP, or data science experts. Tel Delbari, the API team leader at AI21 Labs, expressed his excitement about offering this technology as a hassle-free solution that can be implemented immediately.
The advent of ChatGPT has prompted businesses of all sizes to seek ways to incorporate LLMs into their data stack. The goal is to provide internal teams and customers with faster and more seamless access to accurate and useful information. Traditionally, fine-tuning existing models to suit specific enterprise scenarios has been the approach. However, this method often demands significant engineering resources, making it inaccessible to many companies.
The Contextual Answers API by AI21 Labs eliminates this barrier by enabling businesses to bring any generative AI use case to life right from the start. Uploading documents to the AI21 Labs Studio using the web GUI or API and SDK is the initial step for enterprises to get started. After loading the files, users can directly send questions via the API and receive answers. The API is user-friendly, allowing developers, even those without expertise in NLP or AI, to utilize it effectively.
Once the AI engine is operational, business customers and internal employees can pose free-form questions related to internal support, policy checks, or information retrieval from large documents and manuals. The model leverages the context within the uploaded knowledge base to provide concise and accurate answers. It seamlessly handles both structured and unstructured information.
Delbari emphasized that the model is specifically optimized to adapt to internal jargon, acronyms, project names, and other organizational-specific terms. As long as the documents contain the necessary information, the model automatically learns and comprehends it. Importantly, the model remains grounded in and true to the organizational data and internal language, without mixing in external knowledge or internet-derived information.
In terms of data access and security, the AI engine supports unlimited upload of internal corporate data while ensuring appropriate control measures. Access control and role-based content separation can be enforced by limiting the model’s usage to specific files, a set number of files, a designated folder, or tags and metadata. To guarantee data confidentiality and security, AI21 Labs’ AI21 Studio provides a trusted and SOC-2 certified environment. The company’s commitment to maintaining a separate and protected environment is already recognized by various industries, including banking and pharmaceutical companies. Additionally, the AI engine is compatible with AWS Sagemaker Jumpstart and AWS Bedrock, empowering enterprises to utilize this product’s core capabilities within their virtual private clouds (VPCs).
Looking ahead, AI21 Labs plans to integrate this feature into its writing platform, Wordtune. This integration will enable users to swiftly retrieve specific information from uploaded documents, further enhancing the platform’s functionality and usability.
Leading players in the data ecosystem, Databricks and Snowflake, are also actively working on similar projects. Databricks recently unveiled LakehouseIQ, which utilizes LLM technology to provide context-specific answers to queries on lakehouse data. Similarly, Snowflake has introduced Document AI, a purpose-built multimodal LLM that extracts insights from unstructured documents.
AI21 Labs’ Contextual Answers represents a significant leap forward in leveraging AI technology to unlock the value of data assets. By offering a user-friendly and ready-to-use API, enterprises can effortlessly incorporate LLM technology into their operations, enabling seamless information retrieval and enhanced user experiences. With the integration of Contextual Answers into Wordtune on the horizon, AI21 Labs continues to push the boundaries of AI-powered solutions for data-driven enterprises.
Leave a Reply