The hurdles include data management, security concerns, and limitations in compute resources. However, startups like Piensoare emerging to address these obstacles and simplify AI deployment.
As companies increasingly turn to AI for transformative solutions, the challenges of deploying AI projects persist. A 2023 S&P Global survey reveals that nearly half of companies with AI projects in production remain at the pilot or proof-of-concept stages.
Origins at MIT: Pienso’s Unique Approach
Founded in 2016 by MIT alumni Birago Jones and Karthik Dinakar, Pienso focuses on democratizing AI by offering a platform that enables users to build and deploy models without the need for coding expertise.
![Pienso secures $10 million to builds no-code tools for training AI models image 52](https://i0.wp.com/nosisnews.com/wp-content/uploads/2024/03/image-52.png?resize=1024%2C464&ssl=1)
![Pienso secures $10 million to builds no-code tools for training AI models image 52](https://i0.wp.com/nosisnews.com/wp-content/uploads/2024/03/image-52.png?resize=1024%2C464&ssl=1)
The startup recently secured $10 million in a Series A funding round led by Latimer Ventures, with participation from Gideon Capital, SRI, Uncork, and Good Growth Capital.
Jones and Dinakar’s journey began at MIT’s Media Lab, where they collaborated on a tool to aid social media platforms in moderating and identifying bullying content. The pivotal realization came when they recognized the importance of subject-matter experts, such as teenagers, in training AI models effectively. Pienso was born out of their mission to commercialize tools that empower non-technical professionals.
Democratizing AI for Non-Technical Talent
Pienso positions itself as an AI suite tailored for “non-technical talent,” catering to researchers, marketers, and customer support teams with access to extensive data for AI training but lacking the resources for structuring and analysis.
Jones emphasizes the need to scale AI beyond large language models, stating that Pienso believes any domain expert, not just an AI engineer, should be capable of training and fine-tuning models.
The platform guides users through annotating or labeling training data for pre-tuned open source or custom AI models. This process is crucial for teaching AI to perform tasks by associating labels with specific data, such as pairing an image of a bird with the label “finch.” Pienso’s cloud or on-premises deployment integrates seamlessly with enterprise systems through APIs, ensuring flexibility and security.
Notable Users and Business Model
Notable users of Pienso include the U.K. broadcaster Sky, employing the platform to analyze customer service calls, and an undisclosed U.S. government agency testing it for illegal weapons tracking.
Pienso’s no-code interface enables teams to train models directly using their company’s data, addressing privacy concerns and capturing the unique nuances of each organization.
The startup’s business model involves a yearly license fee based on the number of AI models deployed by companies. This approach allows customers to experiment with building new models before making substantial investments, aligning with Pienso’s commitment to democratizing AI.