During the past decade, while employed at Coursera and later Google, Jiquan Ngiam frequently identified opportunities to automate engineering functions supporting non-developers. Unfortunately, resource constraints always hindered these endeavors. In collaboration with five friends, including his former Google colleague Vijay Vasudevan, Ngiam explored the potential of artificial intelligence models, particularly in cloud computing, for tasks such as reasoning, planning, and coding.
The realization that these models could generate code and enhance reasoning capabilities sparked the idea of creating specialized assistants for non-engineers. This led to the establishment of Lutra AI, a startup founded in April, focusing on creating AI workflows from natural language, eliminating the need for technical expertise. Lutra seamlessly integrates with popular applications like Google Workspaces and Slack, enabling automation for tasks such as email management and internet research.
While numerous companies, including Respell, Unity, Parabola, and even industry giants like Nvidia, are delving into AI workflow solutions, Lutra stands out in several ways. Firstly, Lutra adopts a code-first approach, prioritizing security and reliability during AI workflow execution to safeguard data. Secondly, instead of using large language models (LLMs) for all purposes, Lutra concentrates LLMs on specific tasks to achieve more effective results.
Recently emerging from stealth mode after securing $3.8 million in seed funding from investors like Coatue Ventures, Hustle Fund, Maven Ventures, WVV Capital, and notable individuals such as Andrej Karpathy, Jeff Dean, and Scott Belsky, Lutra is currently in private beta with a limited customer base. Ngiam aims to expand Lutra’s user base and focus on product development and market fit with the new funding.
In terms of best practices for developing a generative AI copilot for business, Ngiam emphasizes the significance of integrating tools across various software applications. Recognizing the pervasive use of technology in contemporary business operations, he envisions empowering companies to operate more efficiently by providing tools that seamlessly integrate with diverse software ecosystems.
