Join our daily and weekly newsletters for the latest updates and exclusive content on industry-leading AI coverage. Learn More
As the demand for AI continues to grow, new tools are emerging to help with development and deployment. One such tool is RagaAI, a California-based startup that recently came out of stealth mode with $4.7 million in seed funding from pi Ventures. Other investors include Anorak Ventures, TenOneTen Ventures, Arka Ventures, Mana Ventures, and Exfinity Venture Partners.
RagaAI, founded by former Nvidia executive Gaurav Agarwal, plans to use the capital to further research and enhance its automated testing platform to ensure safe and reliable AI. Agarwal highlighted that the company is already working with Fortune 500 companies to address issues like bias, accuracy, and hallucinations in various use cases.
Building and deploying AI is challenging. Teams must collect data, train models, and monitor their performance to ensure they meet expectations. Any small error can lead to high costs and missed opportunities. Agarwal experienced these challenges firsthand at Nvidia and Indian mobility company Ola. This inspired him to create an automated testing platform that detects, diagnoses, and fixes AI issues on the fly. RagaAI’s platform performs around 300 tests, identifying a wide range of problems, from data and model issues to operational gaps.
When the platform identifies an issue, it helps users trace it to its root cause, which could be bias in training data, poor labeling, data drift, or lack of model robustness. It then offers actionable recommendations, such as removing poorly labeled data points or retraining the model to address data and concept drift issues.
At the core of RagaAI’s technology are RagaDNA foundation models, which generate high-quality embeddings – compressed and meaningful data representations. These embeddings are used for issue detection, diagnosis, and remediation. Jigar Gupta, the head of product at RagaAI, explained that RagaDNA models are custom-trained for testing purposes, adding intelligence to testing workflows and identifying edge cases or poor-quality training data.
RagaAI’s platform has already made a significant impact. Several Fortune 500 companies, including AI-first companies like LightMetrics and SatSure, are using the technology. For example, an e-commerce company reduced errors in its chatbot, and an automotive company improved its model’s accuracy in detecting vehicles in low-light conditions.
RagaAI believes its technology can reduce 90% of the risks in AI development and accelerate time to production by over three times. With the new funding, the company plans to advance its research and development, improve its platform, expand its team, and raise awareness about the importance of developing safe and transparent AI.
However, RagaAI is not alone in this field. Over the past year, several companies, including Arize’s Pheonix open-source library, Context AI, and Braintrust Data, have emerged with similar goals. Observability players like Acceldata are also focusing on generative AI monitoring to assist with deployment.
As AI is projected to become a $2 trillion market by 2030, the demand for tools ensuring AI safety and reliability is expected to grow significantly. RagaAI believes that up to 25% of this market will be dedicated to such tools.