AI memory startup focused on cutting token costs raises $98 million

• Funding round led by General Catalyst, Kleiner Perkins, and Sequoia, with participation from OpenAI co-founder Andrej Karpathy
• Engram's 'learned memory' approach recalls organization-specific workflows and context to deliver cheaper output, addressing rising costs as new AI models prove more expensive than previous iterations
• The 13-person company, founded in October, plans to use funding for compute resources and talent acquisition

Engram Raises $98M to Slash AI Token Costs Through Memory Technology

Key Investment Details:

AI startup Engram secured $98 million in funding from General Catalyst, Kleiner Perkins, Sequoia, and OpenAI co-founder Andrej Karpathy. The 8-month-old company, founded in October with just 13 employees, positions itself as a solution to rising AI costs affecting corporate developers.

Core Technology:

Engram specializes in "learned memory" for AI systems, enabling models to recall organization-specific workflows and context. The company claims its technology can match or outperform frontier AI labs while using up to 100 times fewer tokens—the fundamental unit of cost for AI queries. This efficiency gain addresses a critical pain point as new, sophisticated AI models prove increasingly expensive, challenging assumptions that scale would reduce costs.

Market Position:

Despite being less than a year old, Engram has secured notable clients including Microsoft, Notion, and legal AI startup Harvey. The company's approach differs from major players like OpenAI and Anthropic by specializing in contextual memory rather than broad capabilities.

Business Rationale:

Founded by CEO Dan Biderman, a computational neuroscience PhD from Columbia who worked at Stanford's AI lab, Engram addresses what he calls the "genius stranger model"—AI that appears smart but has limited memory. The startup aims to build an "intuition layer" that current models lack, helping organizations manage exploding data volumes and costs.

Funding Use:

The capital will support compute infrastructure and talent acquisition as Engram scales its operations to meet growing demand from enterprises seeking to control AI expenditures while maintaining performance.