Guys I am back on @X ...
From tomorrow onwards will post daily about what I am learning these days and what I am working on .
So stay connected. ๐ #connect
Guys I am back on @X ...
From tomorrow onwards will post daily about what I am learning these days and what I am working on .
So stay connected. ๐ #connect
> What it is ?
SentinelSOL makes risk a first-class, verifiable on-chain primitive on Solana. A decentralized network of AI oracle nodes stakes SOL and publishes live risk scores for wallets, tokens, programs and transactions directly into on-chain feed accounts which any protocol can read natively, exactly the way it reads a Pyth price feed (CPI-free, one account read). Scores are set by stake-weighted consensus; nodes that deviate materially from consensus are slashed, and accurate nodes build on-chain reputation. Every feed, node, submission and score is a public Solana account, independently verifiable on Solana Explorer.
> Why it matters ?
Chainlink and Pyth decentralized price data. Risk data is still centralized: Webacy, Chainalysis and similar services sell closed APIs you must trust blindly. No decentralized AI risk oracle exists on any chain. SentinelSOL decentralizes what centralized risk APIs sell with crypto-economic guarantees (staking, slashing, reputation) instead of a brand promise.
> How it works ?
1. Anyone submits a subject (wallet / token / program / transaction) for analysis.
2. Each staked oracle node independently gathers on-chain evidence and runs AI risk analysis (Groq).
3. Nodes publish their scores on-chain; the program computes a stake-weighted consensus score with confidence.
4. Nodes deviating >30 points from consensus are slashed; agreement builds reputation.
5. Protocols read the resulting feed account like a price feed and gate actions on it (e.g. block a borrow if counterparty score > 70).
Iโve been inactive on X because my team and I have been building something exciting.
Launching soon... ๐
We didn't get the funding, but we did get into NVIDIA Inception! ๐ค๐
We'll take the GPUs, cloud credits, and AI resources, and we will keep building. @nvidia Thanks.โค๏ธ
Hey @forwarddeploy sir, I am building something in VoiceAI space.
So anything possible for early stage Startups as credit? @xai@Vapi_AI
As of now I am using @DeepgramAI for voice and @deepseek_ai for the pipeline.
And if you want i would love to give more details about it. ๐