
She | Exploring AI | Web3 | Onchain Culture Content writer | Data analyst | Creator on @base | @Farcaster_xy | Remote Jobs
Earth
Joined October 2024
- Tweets44.1K
- Following3K
- Followers3.9K
- Likes46.6K
Mercor funds have landed! 🥹
Project Odin actually came through 🤭 Two days’ work!
Thank you, Jesus! 💃🏼🙏
Bring your clients! Let's win on WORKADA.
✅ Account creation
✅ Application support
✅ Interview preparation
Get instant interview invites
I'm fully active.
Ouuu 😌🤭 got invited for interview immediately!
DM for Workada account creation, calibration, and account management.
You’re not paying a dime btw.
So let’s run this up, fellas!
Countries eligible: Canada, Ohio, North Carolina, Wisconsin.
1
5
666
Ouuu 😌🤭 got invited for interview immediately!
DM for Workada account creation, calibration, and account management.
You’re not paying a dime btw.
So let’s run this up, fellas!
Countries eligible: Canada, Ohio, North Carolina, Wisconsin.
Another opportunity on Mercor!
The Generalist project is offering $20–$80 per task.
Interested in applying? Use the link below to get started 👇🏽
t.mercor.com/pxfhY
Over the past few posts, I’ve explored different parts of what @primus_labs is building, and one thing that has stood out to me is how these technologies address different challenges around data privacy and verification.
From zkTLS and DVC to confidential payroll, private DeFi, ZK identity verification and AutoHoG, each solution has its own role, but they all contribute to a bigger picture.
What I find interesting is the idea that applications can verify information and perform computations without users having to expose sensitive data unnecessarily.
That’s an important direction for Web3, especially as more real-world activity moves onchain.
There’s still a lot to learn about how these solutions will develop and gain adoption, and I’ll be following primus labs to see what comes next.
GM everyone!
Just so you know, you can earn up to $400 or £400 weekly with USA 🇺🇸 and Spain 🇪🇸 Prolific accounts.
Spain accounts usually come with 40/40 studies upon creation, while USA accounts typically come with 50/50 or 27/27, depending on study availability.
Interested in a waitlist bypass? Come through. Let’s work!
December is just around the corner. Don’t wait until it’s too late.
Over the past few posts, I’ve explored different parts of what @primus_labs is building, and one thing that has stood out to me is how these technologies address different challenges around data privacy and verification.
From zkTLS and DVC to confidential payroll, private DeFi, ZK identity verification and AutoHoG, each solution has its own role, but they all contribute to a bigger picture.
What I find interesting is the idea that applications can verify information and perform computations without users having to expose sensitive data unnecessarily.
That’s an important direction for Web3, especially as more real-world activity moves onchain.
There’s still a lot to learn about how these solutions will develop and gain adoption, and I’ll be following primus labs to see what comes next.
AnitaJude 🌸 retweeted
Over the past few posts, I’ve explored different parts of what @primus_labs is building, and one thing that has stood out to me is how these technologies address different challenges around data privacy and verification.
From zkTLS and DVC to confidential payroll, private DeFi, ZK identity verification and AutoHoG, each solution has its own role, but they all contribute to a bigger picture.
What I find interesting is the idea that applications can verify information and perform computations without users having to expose sensitive data unnecessarily.
That’s an important direction for Web3, especially as more real-world activity moves onchain.
There’s still a lot to learn about how these solutions will develop and gain adoption, and I’ll be following primus labs to see what comes next.
🤝 Paid partnership
AutoHoG by @primus_labs tackles one of the biggest bottlenecks in Fully Homomorphic Encryption (FHE): computational overhead.
FHE allows operations on ciphertexts without decrypting the underlying data. For confidential DeFi, balances and transaction amounts can remain encrypted throughout computation.
The challenge is execution efficiency. Conventional gate-level FHE evaluation involves expensive bootstrapping operations that increase latency and limit throughput.
AutoHoG addresses this at the circuit-compilation layer. It takes standard Verilog circuits through logic synthesis, identifies optimization opportunities, and compiles them into compound homomorphic gates (HoGs).
By grouping compatible operations instead of evaluating each Boolean operation independently, it reduces expensive homomorphic evaluations.
Primus reports:
• Up to 5.7× latency reduction.
• Approximately 90ms for a confidential transfer involving encrypted 64-bit balances.
• Over 210 transfers per second using eight NVIDIA RTX 5090 GPUs.
Its GPU implementation also uses the Number Theoretic Transform (NTT) for exact integer arithmetic and deterministic computation.
AutoHoG targets the execution bottleneck in FHE infrastructure, helping make encrypted computation more practical for real onchain transaction workloads.
AnitaJude 🌸 retweeted
There is a difference between proving that private transactions are possible and making them practical enough for everyday financial activity.
That has been a difficult challenge for FHE.
What @primus_labs is showing with AutoHoG deserves attention because the improvement isn't coming from simply throwing more expensive hardware at the problem. They're changing how encrypted computations are designed and executed.
A 64-bit confidential transfer taking around 90ms on their stack makes the conversation around onchain privacy much more concrete.
Privacy shouldn't force financial applications to compromise on execution speed. That's the gap Primus is working to close.
🤝 Paid partnership
Over the past few posts, I’ve explored different parts of what @primus_labs is building, and one thing that has stood out to me is how these technologies address different challenges around data privacy and verification.
From zkTLS and DVC to confidential payroll, private DeFi, ZK identity verification and AutoHoG, each solution has its own role, but they all contribute to a bigger picture.
What I find interesting is the idea that applications can verify information and perform computations without users having to expose sensitive data unnecessarily.
That’s an important direction for Web3, especially as more real-world activity moves onchain.
There’s still a lot to learn about how these solutions will develop and gain adoption, and I’ll be following primus labs to see what comes next.
🤝 Paid partnership
AutoHoG by @primus_labs tackles one of the biggest bottlenecks in Fully Homomorphic Encryption (FHE): computational overhead.
FHE allows operations on ciphertexts without decrypting the underlying data. For confidential DeFi, balances and transaction amounts can remain encrypted throughout computation.
The challenge is execution efficiency. Conventional gate-level FHE evaluation involves expensive bootstrapping operations that increase latency and limit throughput.
AutoHoG addresses this at the circuit-compilation layer. It takes standard Verilog circuits through logic synthesis, identifies optimization opportunities, and compiles them into compound homomorphic gates (HoGs).
By grouping compatible operations instead of evaluating each Boolean operation independently, it reduces expensive homomorphic evaluations.
Primus reports:
• Up to 5.7× latency reduction.
• Approximately 90ms for a confidential transfer involving encrypted 64-bit balances.
• Over 210 transfers per second using eight NVIDIA RTX 5090 GPUs.
Its GPU implementation also uses the Number Theoretic Transform (NTT) for exact integer arithmetic and deterministic computation.
AutoHoG targets the execution bottleneck in FHE infrastructure, helping make encrypted computation more practical for real onchain transaction workloads.
Mercor funds are finally on their way to me.
What a lovely feeling.
Project Odin did 🤭
Just a two-day work on account 2.
Please bring back Odin🥹
Another interesting part of @primus_labs is its Data Verification and Computation (DVC) workflow.
The idea is to separate data verification from business logic.
Primus uses zkTLS to verify the authenticity of HTTPS requests and responses, without handling the application’s actual computation.
Here’s how the workflow works:
>A dApp submits a verification task to the Primus Network.
>Attestor nodes perform the zkTLS process and return signed attestation data.
>The SDK extracts the required response data locally.
>The attestation and data are passed to a verification program running in a TEE with a zkVM.
>The program checks the attestation, validates the source and response, then executes the application’s business logic.
>The zkVM returns a proof to the dApp.
This separation means Primus handles data authenticity, while developers retain control over the computation their applications need.
It’s an interesting approach to making offchain data useful in onchain applications without relying on trust alone.
AnitaJude 🌸 retweeted
Game creation is one area where AI could make a real difference for independent creators.
Think about someone who has a unique game concept but has never written a line of code. Until now, turning that idea into something playable could require learning development tools or finding someone to build it.
Now, the process can start with a simple description.
The new Agent on Agent.family, powered by claudeai Fable 5.2, gives users a way to experiment with their own game ideas and see them take shape.
I see this as a meaningful use of AI because it gives creativity a more direct path to execution.
@termix_ai is exploring how agents can move into everyday creative tasks and game creation is a good place to demonstrate that potential.
🤝 Paid partnership
🎮 One sentence. One game.
A game creation Agent powered by @claudeai Fable 5.2 is now live on Agent.family!
Tell the Agent what game you want to play and let it bring your idea to life.
Try it now 👇
agent.family/listing?id=cmwd…
Another interesting part of @primus_labs is its Data Verification and Computation (DVC) workflow.
The idea is to separate data verification from business logic.
Primus uses zkTLS to verify the authenticity of HTTPS requests and responses, without handling the application’s actual computation.
Here’s how the workflow works:
>A dApp submits a verification task to the Primus Network.
>Attestor nodes perform the zkTLS process and return signed attestation data.
>The SDK extracts the required response data locally.
>The attestation and data are passed to a verification program running in a TEE with a zkVM.
>The program checks the attestation, validates the source and response, then executes the application’s business logic.
>The zkVM returns a proof to the dApp.
This separation means Primus handles data authenticity, while developers retain control over the computation their applications need.
It’s an interesting approach to making offchain data useful in onchain applications without relying on trust alone.
AutoHoG by @primus_labs tackles one of the biggest bottlenecks in Fully Homomorphic Encryption (FHE): computational overhead.
FHE allows operations on ciphertexts without decrypting the underlying data. For confidential DeFi, balances and transaction amounts can remain encrypted throughout computation.
The challenge is execution efficiency. Conventional gate-level FHE evaluation involves expensive bootstrapping operations that increase latency and limit throughput.
AutoHoG addresses this at the circuit-compilation layer. It takes standard Verilog circuits through logic synthesis, identifies optimization opportunities, and compiles them into compound homomorphic gates (HoGs).
By grouping compatible operations instead of evaluating each Boolean operation independently, it reduces expensive homomorphic evaluations.
Primus reports:
• Up to 5.7× latency reduction.
• Approximately 90ms for a confidential transfer involving encrypted 64-bit balances.
• Over 210 transfers per second using eight NVIDIA RTX 5090 GPUs.
Its GPU implementation also uses the Number Theoretic Transform (NTT) for exact integer arithmetic and deterministic computation.
AutoHoG targets the execution bottleneck in FHE infrastructure, helping make encrypted computation more practical for real onchain transaction workloads.
AnitaJude 🌸 retweeted
AutoHoG by @primus_labs tackles one of the biggest bottlenecks in Fully Homomorphic Encryption (FHE): computational overhead.
FHE allows operations on ciphertexts without decrypting the underlying data. For confidential DeFi, balances and transaction amounts can remain encrypted throughout computation.
The challenge is execution efficiency. Conventional gate-level FHE evaluation involves expensive bootstrapping operations that increase latency and limit throughput.
AutoHoG addresses this at the circuit-compilation layer. It takes standard Verilog circuits through logic synthesis, identifies optimization opportunities, and compiles them into compound homomorphic gates (HoGs).
By grouping compatible operations instead of evaluating each Boolean operation independently, it reduces expensive homomorphic evaluations.
Primus reports:
• Up to 5.7× latency reduction.
• Approximately 90ms for a confidential transfer involving encrypted 64-bit balances.
• Over 210 transfers per second using eight NVIDIA RTX 5090 GPUs.
Its GPU implementation also uses the Number Theoretic Transform (NTT) for exact integer arithmetic and deterministic computation.
AutoHoG targets the execution bottleneck in FHE infrastructure, helping make encrypted computation more practical for real onchain transaction workloads.
Interesting how @primus_labs is connecting Web2 identity data with onchain verification, and the BNB ZK ID SDK is a good example.
It combines Primus zkTLS with Brevis zkVM, giving developers a unified flow to generate proofs from supported platforms like Binance, OKX, GitHub, Steam and Amazon.
Here’s how the pipeline works:
>The dApp initializes the SDK with its registered appId.
>It selects the required identityPropertyId.
>Primus zkTLS verifies the relevant data from the source platform.
>The proof request is submitted to the BNB ZK ID Gateway.
>Brevis zkVM handles the next proving stage, and the result is returned as an onchain attestation.
The important part is that an application can verify a specific identity property without requiring the user to expose their entire account history.
It’s a useful bridge between offchain identity data and onchain applications.
Everyone talks about the entry. Nobody wants to talk about the exit. That’s where the real asymmetry lives, and it’s worth looking at how both platforms handle the moment you want out.
This is why @hypesium and @RebateLaunch are worth reading together.
@RebateLaunch locks you in with time. If you buy a token and hold it, your share of the trading fees climbs from 1x to 2x over seven days. But the moment you sell, that multiplier resets. You aren't just paying a sell tax; you're forfeiting the higher payout tier you spent a week building. The contract makes leaving expensive.
@hypesium handles the exit differently. It doesn't lock you in. Instead, it simulates the exit before you even enter. Before the bot fires your buy, it runs a simulation of the sell route. It checks the tax cap and the liquidity floor. If the sell test fails, it doesn't send the buy. It assumes the exit might not exist when you need it, and it tries to prove that before your funds are at risk.
One platform makes the exit expensive to discourage leaving. The other makes the exit visible to prevent you from entering a trap.
I see three links in the chain:
The incentive: Rebate pays you to stay, while Hypesium just tries to make sure you can leave.
The risk: Rebate’s model relies on the fee pool staying full. Hypesium’s model relies on the simulation being accurate.
The reality: Neither can force a buyer to appear when you want to sell.
A locked contract is not the same as a liquid market. A perfect simulation is not the same as a guaranteed fill. Before leaning on either story, I'd check:
How much of Rebate's payout pool is funded by organic volume versus the initial launch spike?
Does Hypesium’s sell simulation account for rapid price impact, or just honeypot detection?
What does the exit actually look like on day 30, not day one?
A mechanism that works during the launch week is not the same as an economy that lasts. The mechanics are designed to look good on paper. Usage is what shows if the design holds up.
When a platform makes the exit expensive, which metric would you trust first: the contract design, the current payout rate, or the depth of the order book?
AnitaJude 🌸 retweeted
Most Web3 projects talk about transparency. But how easy is it to actually verify what’s happening? 👀
This is why PrimusScope caught my attention. 🚀
An independently built analytics platform for @primus_labs that brings together:
🏦 Finance metrics — NAV, APY, deposits & flows
📊 zkTLS network activity & attestor performance
🔎 On-chain task & transaction explorer
🛠 APIs, CSV exports & developer tools
📈 Custom dashboards & saved research
No wallet required. Just public data you can explore and verify.
This is exactly the kind of community-built infrastructure Web3 needs.
🔥 Check out the 90-second demo!
🔗 App + Kaito campaign links in the thread below 👇
🚀 I built PrimusScope - an independent analytics platform for the @primus_labs ecosystem.
One place to explore real activity, verify public data, and understand what’s happening across the ecosystem.
🎬 Here’s what I built, in a 90-second walkthrough 👇
🏦 Finance Analytics
Explore U/eU and USDT/eUSDT routes. Track public batch activity, strategy assets, shares, NAV, APY, deposits, withdrawals, and net flows.
📊 zkTLS Network Analytics
Monitor requests, attestor reports, response coverage, and latency over time. Filter by events, blocks, or individual tasks.
🔎 Transaction & Task Explorer
Paste a task or transaction ID to investigate its timeline, inspect recorded events, and verify activity through original on-chain links.
🛠 Developer Tools
Query public data through read-only APIs, inspect JSON responses, export CSV datasets, and access official SDK documentation.
🔖 Personal Workspace
Save dashboards, research periods, and individual tasks so you can return to your analysis anytime.
🧾 Transparent Data
Every metric comes with source references, timestamps, and coverage details. Shared-strategy analytics are clearly distinguished from confidential Primus positions.
Why does PrimusScope matter?
👥 Users can understand what’s actually happening instead of relying on announcements.
👨💻 Builders can explore real network activity, access structured data, and create their own tools.
🌐 The Primus team gets an independent, community-built analytics resource to understand adoption, network behavior, and developer needs.
The goal is simple: make public Primus data useful, accessible, and verifiable.
🔐 Private data. Public insight.
🎥 Watch the 90-second walkthrough and see PrimusScope in action.
🔗 Try PrimusScope - link in the reply below.
@primus_labs @xxiang_xie - I’d love your feedback!
What would make PrimusScope more valuable to your team and ecosystem builders?
I’d especially appreciate input from the Primus DevRel team on analytics, API capabilities, and developer workflows.
Built for the Primus community. Open to making it better together. 🚀
AutoHoG by @primus_labs tackles one of the biggest bottlenecks in Fully Homomorphic Encryption (FHE): computational overhead.
FHE allows operations on ciphertexts without decrypting the underlying data. For confidential DeFi, balances and transaction amounts can remain encrypted throughout computation.
The challenge is execution efficiency. Conventional gate-level FHE evaluation involves expensive bootstrapping operations that increase latency and limit throughput.
AutoHoG addresses this at the circuit-compilation layer. It takes standard Verilog circuits through logic synthesis, identifies optimization opportunities, and compiles them into compound homomorphic gates (HoGs).
By grouping compatible operations instead of evaluating each Boolean operation independently, it reduces expensive homomorphic evaluations.
Primus reports:
• Up to 5.7× latency reduction.
• Approximately 90ms for a confidential transfer involving encrypted 64-bit balances.
• Over 210 transfers per second using eight NVIDIA RTX 5090 GPUs.
Its GPU implementation also uses the Number Theoretic Transform (NTT) for exact integer arithmetic and deterministic computation.
AutoHoG targets the execution bottleneck in FHE infrastructure, helping make encrypted computation more practical for real onchain transaction workloads.
Interesting how @primus_labs is connecting Web2 identity data with onchain verification, and the BNB ZK ID SDK is a good example.
It combines Primus zkTLS with Brevis zkVM, giving developers a unified flow to generate proofs from supported platforms like Binance, OKX, GitHub, Steam and Amazon.
Here’s how the pipeline works:
>The dApp initializes the SDK with its registered appId.
>It selects the required identityPropertyId.
>Primus zkTLS verifies the relevant data from the source platform.
>The proof request is submitted to the BNB ZK ID Gateway.
>Brevis zkVM handles the next proving stage, and the result is returned as an onchain attestation.
The important part is that an application can verify a specific identity property without requiring the user to expose their entire account history.
It’s a useful bridge between offchain identity data and onchain applications.
AnitaJude 🌸 retweeted
NFT ownership could become more interesting when it comes with actual utility.
With @meetmotes, an NFT can represent more than digital ownership. Each Genesis NFT is connected to an AI agent with its own smart account.
What matters here is that the owner still sets the boundaries. You decide what the agent can spend, which actions it can perform and how much control it gets.
That makes autonomous execution more practical because giving an agent responsibility doesn't mean surrendering control.
So, you people posting that Project Odin is back, Wetin Una gain?
Make Una no rage bait me, I Dey quick scatter things.🌚
Busy preparing everyone for detty December! 🤭
I am actively creating accounts and by bypassing Prolific wait-list
You can skip the Prolific wait-list by following my steps properly.
I have rules.
Go more for USA 🇺🇸 accounts for more studies.
Come for your wait-list bypass.
Today appears favourable for UK 🇬🇧 Prolific accounts.
Studies offering higher pay are dropping.
Come for wait-list bypass and start earning.
2
376





























