$OCT seems like an asymmetric privacy bet that's under the radar. @octra is a FHE L1 that lets the network compute on encrypted data without ever decrypting it. They're built around a proprietary HFHE (Hypergraph FHE) which enables private AI models, confidential DEFI, and encrypted agents with no TEEs, no threshold decryption, and no trusted hardware. Most of the other FHE projects such as ZAMA, Fhenix, Inco, etc. are basically co-processors or libraries that are bolted on existing chains and often rely on threshold decryption, whereas Octra is a native L1 that has its own HFHE scheme from scratch, open-sourced the core PoC + benchmarks, and rejects threshold decryption entirely. They've raised roughly $8m (pre-seed + Echo), and the pre-seed was led by investors like Finality Capital, Big Brain, Karataga, Presto Labs, etc. They were also selected by Cobie's Echo consecutively and both rounds were oversubscribed and were approved into the first cohort of Sonar. They were supposed to have a $20m Sonar launch at $200M FDV, but later pivoted to a fair on-chain Uniswap CCA with no extreme VC concentration. Team has been building for 5+ years and were talking about "FHE" before it became mainstream. They didn't try to extract with high FDVs and roughly 35% of the supply is burnt via on-chain governance. It's a serious FHE L1 play focused on private AI + general encrypted computation, and they've shipped far more than their competitors, who are valued over 10x more. Mainnet alpha is live and dev has said that the recent lite node update was the penultimate pre-release update before mainnet, and with a chart that's -70% from ATH, this set up looks too good to ignore.