Trajectory raises $45M to make AI model training accessible to nontechnical experts
techcrunch ·
Trajectory, a startup founded by three Stanford classmates with backgrounds at Google DeepMind and Apple, has raised $45 million in a Series A funding round led by Sequoia Capital. The company aims to democratize AI model training by automating the fine-tuning process for non-technical experts through continual learning. This approach allows businesses to connect their applications to Trajectory’s platform, enabling models to learn from feedback and improve at specialized tasks without requiring expensive internal research teams or consultants. Co-founder Ronak Malde compares foundational models to smart PhD students who lack practical experience, whereas Trajectory helps them gain decades of specialized knowledge. Current customers like Harvey and Rogo use the platform to post-train open-source models, enhancing performance in fields such as law and finance. Unlike autonomous self-improvement systems, Trajectory keeps humans involved in the loop to ensure safety and alignment with organizational standards. The startup differentiates itself through its experienced team and a user-friendly interface designed to make post-training accessible to specialists who are not AI engineers.
techcrunch