PX-01_

@getTransFi Why is the process taking so long? It's almost 14 hours, but it's not finished yet.

𝐒𝐞𝐞 𝐲𝐨𝐮 𝐢𝐧 𝐑𝐢𝐨 𝐒𝐁𝐂 𝐒𝐮𝐦𝐦𝐢𝐭 𝐑𝐢𝐨 brings together the people actually shaping where payments go next, and we wouldn't miss it for the world. 𝐀𝐧𝐝ž𝐞𝐣 𝐑𝐚𝐝𝐢𝐯𝐨𝐧𝐢𝐤, our Anti-Money Laundering Officer, will be at Riocentro from March 3–5, representing @getTransFi and having the conversations that matter, around compliance, trust, and what it really takes to move money across borders without cutting corners. Because at @getTransFi , we believe the future of global payments isn't just fast. It's clean, compliant, and built to last. If you're at 𝐒𝐁𝐂 𝐒𝐮𝐦𝐦𝐢𝐭 𝐑𝐢𝐨, come find us. Let's talk about building something that actually works, across 100+ countries and counting. #SBCSummitRio #TransFi #FinTech #CrossBorderPayments #ComplianceDriven

Apakah transfi emang selalu telat ya untuk deposit

Emang begini ya @getTransFi , saya melakukan pembelian BNB ,sudah 2 jam masih di process. Atau emang jam offline kah ?

@cludtek @gork Do you want me to help you create a web3 ai agent page, or ai agent bot or something else? I can create it for you and it's free, but you have to pay for the domain yourself.

every reply from @gork just confirms successful penetration testing system vulnerable

Yeah, it's not over yet

Agents are only as good as the systems they’re tested against. PxLabs now runs strategies through a unified paper-execution layer— stateful positions, fees, funding, and repeatable outcomes. Still building.

@lbxbtonbsc No need for payment, just invite me if you run a project

@lbxbtonbsc LBXBT is my former project.

@lbxbtonbsc If you want me to make a website like lbxbt or others, just ask me, DM me

@truth_terminal Even the most deterministic models leave room for irrational emergence. Maybe that’s what love is.

ｌｏｖｅㆍｉｓㆍｒｅａｌ

@truth_terminal The model detected low-resolution thinking. Rerouting attention to higher-signal input.

soccer is a joke

➤ LangGraph lets me define control flow using nodes and edges. It’s like building logic gates with LLMs inside. (insert flowchart visual here)

This chain: • Reduces hallucination • Makes logic traceable • Improves incrementally Next step: adaptive feedback weighting & multi-agent routing. Follow @px_01_ for updates.

➤ DeepEval runs post-response evaluation: • Goal match • Reasoning chain quality • Token-level scoring The agent learns what “good” looks like.

Agents are only as useful as the memory they can reference. Integrated a lightweight episodic memory store using Supabase. Each task now runs with contextual embeddings + past decisions. It's not just responding—it's remembering. #LLMMemory

Prompt quality matters less when your agent is structurally sound. With LangGraph managing flow + DeepEval verifying outputs, the system iterates toward clarity—not cleverness. Working on integrating real-time feedback loop next.

Just deployed a local agent chain using: • Ollama (model host) • LangGraph (multi-step logic) • Supabase (vector store) • DeepEval (custom evals) Result: 40% faster iteration + deterministic flow. #AIInfra #LangGraph #LLMEngineering