Join our daily and weekly newsletters for the latest updates and exclusive content on industry-leading AI coverage.
In today’s fast-paced business world, everyone is looking to leverage AI for better results. Vendors are in a rush to roll out new tools that help customers build top-notch AI/ML applications easily. While many have focused on low-code development, a California-based company called Iterate is trying to eliminate the need for coding altogether. They have just launched AppCoder LLM, a specialized model that instantly produces updated, functional code for AI applications using natural language prompts.
Integrated into Iterate’s Interplay application development platform, AppCoder LLM takes text prompts and generates code much more effectively than existing solutions like Wizardcoder. This tool gives developer teams quick and accurate code for a variety of AI needs, whether it’s object detection or document processing.
This innovative model accelerates the development process by generating workable code quickly. Developer teams are encouraged to try out Interplay-AppCoder LLM to experience the ease of automatic code generation.
Interplay is a fully containerized, drag-and-drop platform that combines AI engines, enterprise data sources, and third-party services to build production-ready applications. With AppCoder, users can simply provide natural language instructions, and the tool generates the required code. Whether working with computer vision libraries like YOLOv8 for advanced object detection or generating code for tools like LangChain and Google libraries, AppCoder makes it possible.
For example, a fast-food drive-thru could connect a video source and ask AppCoder to write a car identification application using the YOLOv8 model. The tool delivers the needed code almost instantly. According to testing, a core detection app was built in just under five minutes, greatly reducing development time and boosting team productivity.
When comparing code-generating LLMs, AppCoder LLM not only delivers faster but also produces better quality outputs than Meta’s Code Llama and Wizardcoder. In specific benchmarks, AppCoder showed significantly higher functional correctness and usefulness scores, indicating its superior ability to run unit tests and produce human-readable, logically ordered code.
Response times for generating code on an A100 GPU are typically between 6-8 seconds. These impressive results come from meticulously fine-tuning various large models on a hand-coded dataset of commonly used generative AI libraries.
Although AppCoder is available now, Iterate sees this as just the beginning. They plan to build 15 private LLMs for large enterprises and aim to expand their models to CPU and edge deployments for better scalability. The company is committed to continuously offering a platform and tools for managing AI engines, new language models, and large datasets, all optimized for rapid development on diverse architectures.
Iterate has seen significant growth in the past two years, nearly doubling its revenue. Their client base includes Fortune 100 companies in banking, insurance, documentation services, entertainment, luxury goods, automotive services, and retail.
Stay informed! Get the latest news delivered to your inbox daily. Subscribe to our newsletter!