Fresh out of Y Combinator, startup atla has announced a successful $5 million Seed funding round, with Creandum leading the investment. The timing coincides with a nearly 24-hour debate among EU lawmakers on the Artificial Intelligence Act, highlighting the relevance and importance of AI regulations.
Generative AI, while powerful, is not without its flaws, as demonstrated by instances of AI-generated images and conversations lacking accuracy and reliability. The limitations of current generative AI technologies become apparent when considering applications in high-stakes sectors such as legal, financial, and medical.
One contributing factor to the challenges faced by generative AI is its training on large language models using textual datasets from sources like Wikipedia and GitHub, which may contain inaccuracies or misinformation.
Addressing these issues is the focus of atla, a member of Y Combinator’s S23 cohort. While specific details about their approach are not provided, atla states that it is developing technology to enhance truthfulness, reduce harmfulness, and increase reliability for generative AI models.
The startup has already conducted pilot programs with legal teams at Volkswagen and N26. Acting as a research assistant, atla’s AI provided generated answers directly cited from reputable legal sources. The successful implementation of atla’s concept demonstrates its potential value in improving the accuracy and reliability of information generated by AI.
Co-founded by Maurice Burger, an AI startup veteran, and Roman Engeler, a researcher specializing in large language models, atla is now poised to scale its offerings and accelerate commercialization. The Seed round was led by Creandum, with participation from existing investor Y Combinator and Rebel Fund, featuring notable partners like Steve Huffman (Reddit), Daniel Kan (Cruise), Max Mullen (Instacart), Trip Adler (Scribd), and Sebastian Mejia (Rappi).
