Connectory raises pre-seed from Bluepoint Partners for AI decision infrastructure
What's the deal? Connectory, a South Korean AI decision-infrastructure startup, has raised a pre-seed round from Bluepoint PartnersDealroom has a profile for this one. Try Dealroom →, it said on July 14, 2026. The amount was not disclosed.
What's the endgame? The company builds "decision infrastructure" that structures information into a decision-ready form before an AI model reasons over it. Its core platform, Atomic Information Decomposition (AID), converts unstructured information into structured intelligence.
Why now? As generative AI spreads, the volume of information to process is rising sharply. Connectory aims to solve a familiar trade-off — adding information raises cost, while cutting it lowers accuracy — by reducing token and compute costs while holding accuracy steady.
The founding team draws on security and intelligence backgrounds. Chief executive officer Lee Ji-won, a former deputy head at S2W, leads business strategy and product design; Chung Jin-woo, who ran AI research at S2W, is CTO. Kim Sun-hee, a former third deputy director at the National Intelligence Service, oversees global policy and regulatory intelligence.
Connectory has completed a proof of concept with a large domestic financial firm and was selected for Shinhan Financial GroupDealroom has a profile for this one. Try Dealroom →'s Shinhan Future's Lab startup program. It plans to expand from finance into semiconductors, aerospace, security, and global policy and regulatory analysis.
What could go wrong? The plan hinges on scaling one approach across industries with very different requirements, from finance to aerospace. Bluepoint senior investment manager Ahn Hwi-jae called the cost-accuracy trade-off "a structural challenge every company will face," citing the team and the market potential as reasons for the investment.
The signal: The bet reflects a shift in AI value away from ever-larger models toward efficiency. As Lee put it, competitiveness in the generative AI era rests "not on a bigger model but on how much more accurate and reliable a decision you can make at the same cost."
Read more: WOW!Tale
