Insider Brief
- Delphina, an AI tool powered by large language models (LLMs) for data science, has secured a $7.5 million seed funding led by Costanoa Ventures and Radical Ventures.
- The investment will support team expansion and product development acceleration.
- The platform aims to transform predictive AI workflows in data science by automating routine tasks, thus reducing project timelines from months to days. It’s designed for enterprises without specialized AI teams, aiding in applications like forecasting and fraud detection.
PRESS RELEASE — SAN FRANCISCO/December 13, 2023 (GLOBE NEWSWIRE) — Delphina, an LLM-powered copilot for data science, today announced the closing of a $7.5 million seed round co-led by Costanoa Ventures and Radical Ventures with participation from 20+ prominent angel investors, including Stanford professor Fei-Fei Li. The funding will be used to expand its team and accelerate product development.
Delphina will do for data science and predictive AI what GitHub Copilot is doing for software engineering. It uses large language models (LLMs) to accelerate the routine, time-consuming tasks in predictive AI workflows, reducing them from months to days. By identifying and preparing relevant data, training models and deploying pipelines, the Delphina platform makes it dramatically faster and simpler for any enterprise — even those without dedicated AI teams — to productize common AI business use cases, like forecasting, personalization, pricing or fraud detection, to name a few.
The goal is to close the gap preventing businesses from using AI in more core business processes. According to Deloitte, 94% of business leaders agree that AI is critical to success over the next five years; however, 41% of business leaders cite slowdowns due to gaps in technical skills. By automating time-consuming tasks, Delphina will enable data science teams to build and deploy better AI models, faster.
“Today, even the most advanced data science teams are mired in painstaking work that AI is better suited to do,” said Jeremy Hermann, Co-founder of Delphina. “With the recent advancements in large language models, it is only now possible to massively improve data science productivity so the world gets the value it wants out of AI faster.”
Many existing tools focus on MLOps (machine learning operations), but the creation and tuning of predictive AI models themselves remains labor-intensive and under-served. The few tools that do target model building favor low-code approaches, whereas Delphina is optimizing for technical builders who code.
“There’s been a boom in AI tools in recent years, but building valuable models and putting them into production has not gotten easier,” said Tony Liu of Costanoa Ventures. “Delphina is taking a very different approach to the problem by using LLMs in a way that was not previously possible. Jeremy and Duncan understand the space better than anyone else.”
“Large language models promise to automate and transform broad swaths of traditional data science and predictive AI workflows, just like we are seeing them do with software engineering,” said Rob Toews of Radical Ventures. “Venture capitalists like to talk about ‘founder/market fit.’ I can’t imagine a stronger founding team than Duncan and Jeremy to tackle this problem, as two top data science leaders who have been at the forefront of this field for years.”
Delphina is deployed with five design partners. They expect full commercial availability in the second half of 2024.
About Delphina:
Delphina is a copilot for data science that speeds predictive AI model building and deployment. Backed by a proven team of data scientists and engineers, the company is led by Co-Founders Jeremy Hermann and Duncan Gilchrist. Delphina was founded in 2023 and is headquartered in San Francisco.
