Improving domain adaptation for language models
%201%20(1).png)
Researchers at UC Berkeley have introduced RAFT (Retrieval-Augmented Fine-Tuning) by leveraging Meta Llama 2 on Azure AI Studio. RAFT enhances domain adaptation in language models by improving their ability to retrieve and integrate relevant information. This novel approach benefits specialized applications by making Meta Llama 2 more versatile and adaptable to various domain-specific tasks.
.png)
46
AI use cases in
Education
.png)

%201%20(1).png)
Rising Academies integrated Claude-powered chatbots into mobile educational tools via WhatsApp to provide both a virtual math tutor (Rori) and a teacher support system (Tari). Rori delivers personalized math lessons through a mix of pre-written lesson plans and dynamic conversations, while Tari offers on-demand curriculum and lesson planning support for teachers. These solutions enhanced personalized learning and professional development in regions with limited access to traditional edtech.
%20(1).png)

%201%20(1).png)
Pensieve integrated Anthropic's Claude models (Claude 3.5 Sonnet and Claude 3.7 Sonnet) to build AI teaching assistants that automate grading and provide 24/7 personalized tutoring. The system clusters student submissions to calibrate grading rubrics and transforms static PDFs into interactive worksheets while enforcing instructor-controlled policies. This integration streamlines course management workflows and enhances real‐time classroom analytics.
%20(1).png)

%201%20(1).png)
Super Teacher integrated Anthropic's Claude to generate initial code for interactive educational tools and draft lesson plans for subjects ranging from pre-K to 5th grade. They embedded the AI into both software development and content creation workflows with mandatory human review to ensure safety and quality standards. This integration streamlined operational processes, enabling engineering and content teams to focus on higher-level development and creative enhancements.
%20(1).png)
62
companies using
Code Agents
.png)

%201%20(1).png)
Bito implemented AI-powered developer agents by integrating Anthropic's Claude and leveraging Claude 3.7 Sonnet for advanced reasoning into its code review and coding workflows. They utilized Anthropic’s robust API and developer-friendly infrastructure to embed an AI Code Review Agent and Bito Wingman directly within developers’ Git workflows and popular IDEs, enabling automated analysis of pull request diffs and code architecture. This integration streamlined code review, error detection, and code generation processes while upholding security standards.
%20(1).png)

%201%20(1).png)
Augment code integrated Anthropic's Claude within Google Cloud's Vertex AI to develop an AI-powered code assistant that provides expert-level contextual understanding of complex software systems. By automating code comprehension, debugging, documentation, and change propagation within development workflows, the solution dramatically reduced project timelines and sped up developer onboarding while ensuring SOC 2 Level 2 compliant security protocols.
%20(1).png)

%201%20(1).png)
Tata Elxsi integrated Microsoft GitHub Copilot into its video distribution platform development and testing processes to provide intelligent code suggestions, optimize existing code, assist in both manual and automated testing, simplify documentation, and enable smoother code translation. The integration streamlined coding and debugging workflows across the development cycle, enhancing productivity and ensuring compliance with strict security standards.
%20(1).png)
49
solutions powered by
Meta
.png)

%201%20(1).png)
Roboflow uses Meta's Segment Anything Model (SAM) to enable users to automatically segment objects in images and videos, significantly reducing the time required to create training datasets for computer vision models.
%20(1).png)

%201%20(1).png)
Untukmu.AI, an online gifting site in Indonesia, uses Meta's Llama 3.1 8B model with split inference processing to protect customer privacy. By running part of the AI model on customers' devices and the rest on their servers, they deliver personalized gift recommendations without accessing or storing personal data. This ensures customer privacy while still providing high-quality, tailored suggestions, enhancing trust and satisfaction.
%20(1).png)

%201%20(1).png)
CodeGPT, a popular coding assistant with over 1.4 million downloads, integrates Meta's Llama models to enhance developer productivity. By using Llama 3.2 (90B), CodeGPT helps developers not just generate code but also answer questions about their codebase, debug code, and onboard new team members. It includes a codebase graph mechanism that lets Llama understand entire repositories, allowing developers to effectively "talk" with their code. This integration leads to at least a 30% increase in productivity and accelerates onboarding from months to days.
%20(1).png)
284
AI use cases in
North America
.png)

%201%20(1).png)
Cox Automotive integrated Claude via Amazon Bedrock into its portfolio by first creating a sandbox environment to evaluate performance metrics and then selecting Claude 3.5 Sonnet for complex tasks and Claude 3.5 Haiku for high-volume content generation. They automated personalized dealer-consumer communications, generated engaging vehicle listing descriptions, and produced SEO-optimized blog posts, while also streamlining internal data governance through automated metadata generation. This integration optimized operational efficiency across marketing and internal data processes.
%20(1).png)

%201%20(1).png)
Intuit integrated Google Cloud’s Document AI and Gemini models into its GenOS platform to automate the autofill of ten common U.S. tax forms, including complex 1099 and 1040 forms. The solution extracts and categorizes data from uploaded documents, drastically reducing manual data entry for TurboTax customers. This integration streamlines tax preparation workflows and improves speed and accuracy.
%20(1).png)

%201%20(1).png)
Block implemented Anthropic’s Claude models (Claude 3.5 Sonnet and Claude 3.7 Sonnet) on its Databricks platform to power its internal AI agent, codename goose. They integrated the LLM using secure OAuth-enabled connections and a custom MCP server to connect internal databases and tools, enabling employees across all roles to auto-generate SQL queries, analyze complex data, and automate workflows. This agentic integration streamlined software development, design prototyping, and data analysis by translating user intents into actionable insights.
%20(1).png)