Observed Signal · Jun 23, 2026 · Funding · Source: CNBC Technology · Impact: 3/5 · Sentiment: Positive
Engram raises $98M to cut AI token costs
Engram, an AI "memory" startup founded in October, announced a $98 million funding round led by investors including General Catalyst, Kleiner Perkins and Sequoia, with participation from OpenAI co‑founder Andrej Karpathy. The 13‑person company says its learned‑memory models store organization‑specific workflows and context to produce accurate responses while using up to 100x fewer tokens than frontier models, aiming to reduce the soaring cost of running modern AI. Engram counts Microsoft, Notion and legal AI startup Harvey among early clients and plans to use the capital to expand compute capacity and hire talent. CEO and co‑founder Dan Biderman described the company’s focus on specialized memory and context rather than general-purpose model capabilities.
A sizeable VC round for an LLM‑adjacent startup focused on reducing token costs could materially affect the economics of deploying large models — lowering operational costs for enterprises and accelerating adoption of AI features across industries including AdTech.
Track General Catalyst Signals & Market Shifts in Real-Time
Polaris7 autonomous intelligence agents track regulatory filings, primary sources, executive changes, and deal flow 24/7. Create your free Explorer workspace to monitor these entities.
Key Takeaways & Evidence Grounding
- Engram raised $98 million in a funding round announced June 23, 2026.
- Investors include General Catalyst, Kleiner Perkins, Sequoia, and Andrej Karpathy.
- Engram claims its models can match or outperform frontier labs while using up to 100 times fewer tokens.
- The company was founded in October and has a 13‑person team.
- Early clients include Microsoft, Notion and legal AI startup Harvey.
Connected Companies & Entities
7 Entities mappedOntology Mapping & Concepts
Related Market Signals & Shifts
Recent verified developments and strategic activity across this market segment.
Memory Costs Surge as AI Infrastructure Complexity Grows
TechCrunch reports that memory (DRAM and cache management) is becoming a central cost and operational factor for running AI models. DRAM prices have risen roughly sevenfold in the past year, and companies are increasingly focused on orchestrating memory so the right data is available to agents at the right time. Anthropic’s prompt-caching pricing (with 5-minute and 1-hour cache windows) illustrates commercial trade-offs: cached reads are much cheaper, but adding data can evict other cached items. Semiconductor analyst Doug O’Laughlin and Val Bercovici (Weka) discuss hardware choices (DRAM vs HBM) and higher-level orchestration. Startups such as Tensormesh are addressing cache optimization. Better memory orchestration and more efficient models can materially reduce token use and inference costs, improving the economics of AI applications.
Etched Raises $700M, Valued at $21B
Etched announced a $700 million funding round that values the AI-hardware startup at $21 billion, led by quant firm Jane Street, which tested and purchased Etched’s hardware. The raise follows a rapid increase in valuation from $5 billion in December and $10.3 billion in July. Etched sells full AI systems it calls "frontier inference clusters" and says it built a low-voltage prefill chip and a new cluster-scale memory/interconnect to accelerate inference (prefill and decode phases), aiming for higher speed and lower cost. Investors include major venture firms such as Kleiner Perkins, Sequoia, Andreessen Horowitz, Tiger Global, Bain Capital Ventures and others.
Memory to the Moon: AI Drives Memory Price Surge
An a16z 'Charts of the Week' analysis documents a sharp, recent run-up in memory prices driven by AI infrastructure demand. DRAM contract prices more than tripled year-over-year by March, while NAND roughly doubled; High Bandwidth Memory (HBM) is being prioritized for model training. The boom has produced outsized profits for memory manufacturers (Samsung, SK Hynix, Micron) and is expected to materially raise operating income in 2026. Capacity expansion lags demand—manufacturers favor higher-margin HBM and hyperscalers are signing multi-year supply contracts—putting pressure on consumer memory supply and potentially raising device prices ~10–20%. The piece also highlights AI-driven productivity gains (Affirm’s AI-first retool increased pull requests and ‘agentic’ output) and broader early-stage adoption of AI agents across industries.
Track Real-Time Market Signals & Shifts
Set up custom watchlists to receive automated, evidence-grounded executive digests whenever material signals or shifts occur across your tracked landscape.