Analysis
Trajectory, a roughly one-year-old AI infrastructure startup, has raised a $45 million Series A led by Sequoia Capital at a $300 million valuation, TechCrunch reported on November 2, 2026. Nvidia, Bessemer Venture Partners and Bain Capital Ventures also participated in the round.
What Trajectory does
Trajectory built a platform for continual learning: companies connect it to their AI applications, it collects real-world performance feedback, and it automatically retrains models for specialized tasks. The company describes the product as automating the job normally done by "forward-deployed engineers" -- the expensive, hands-on technical staff that AI labs and infrastructure vendors like Palantir have used to customize models for individual enterprise customers. Trajectory already counts Harvey, Rogo, Decagon, Mercor and Nvidia itself among its customers.
Who founded Trajectory
Trajectory was founded roughly a year ago by three Stanford classmates -- Ronak Malde, Michael Elabd and Arjun Karanam -- who built a team of 18 people with prior experience at Google DeepMind, Apple, OpenAI, Stripe and Netflix, according to TechCrunch. That pedigree, paired with a customer list that already includes five recognizable enterprise AI names, is a big part of why Sequoia was willing to lead a $45 million round into a company barely a year old.
“Trajectory already counts Harvey, Rogo, Decagon, Mercor and Nvidia itself among its customers.”
Trajectory's funding and valuation
The $45 million Series A values Trajectory at $300 million. The company's investor list mixes a top-tier generalist lead (Sequoia) with a strategic participant that is also a customer (Nvidia) and two growth-stage specialists (Bessemer, Bain Capital Ventures) -- a cap table built for both early product validation and a credible path to a much larger future round.
Competitors in continual learning and post-training infrastructure
Trajectory is entering a crowded and well-capitalized category:
- Thinking Machines Lab -- founded by former OpenAI CTO Mira Murati; working on related post-training infrastructure with substantially more funding and public attention.
- Applied Compute -- valued at $3.25B after a $350M raise just a month before Trajectory's round closed, according to Applied Compute's own announcement -- roughly 10 times Trajectory's valuation.
- Prime Intellect -- a third named competitor working on adjacent decentralized-training infrastructure.
The gap shows how much capital investors are willing to put behind the thesis that continual learning and post-training optimization becomes standard AI infrastructure, a scramble Pulse has also tracked in SWI Compute's $1 billion Nvidia chip order for AI-factory capacity.
The differentiation question
The risk for Trajectory is differentiating at a much smaller scale than its best-funded rivals: Applied Compute's $3.25 billion valuation suggests investors already have a favorite in this category, and Trajectory will need its five current enterprise customers to visibly outperform alternatives -- including simply better prompting or periodic fine-tuning -- to justify a follow-on round anywhere near Applied Compute's trajectory. Whether continual learning proves to be a durable, separately fundable infrastructure layer, or just a feature foundation-model labs eventually build themselves, is the open question this round doesn't answer.