Trung Tran

[Trung Tran @TrungTranlh]

An L1 needs something worth building on.

@ama_protocol launched AMA Hub as a simple gateway for creating privacy-first trading agents using plain language.

No-code agent composition.

Real users.

Real transactions.

[Trung Tran @TrungTranlh]

3/3
Private internals.
Accountable outcomes.

[Trung Tran @TrungTranlh]

PRIVATE DOESN'T MEAN BLIND

���� 1/3
Privacy and transparency are often treated like opposites.
They don't have to be. @ama_protocol

[Trung Tran @TrungTranlh]

2/3
Sensitive strategy can remain confidential while execution produces verifiable results.

[Trung Tran @TrungTranlh]

@ama_protocol AMA just crossed $2M in Economic Agentic Value.

A useful milestone.

But watch the pattern:

mandate → private execution → settlement → proof

That's the infrastructure agents need.

[Trung Tran @TrungTranlh]

3/3
That's the difference between an AI assistant and financial infrastructure.

[Trung Tran @TrungTranlh]

@ama_protocol ��� AI WITH GUARANTEES

🧵 1/3
AI can generate impressive answers.
Financial systems need stronger guarantees.

[Trung Tran @TrungTranlh]

2/3
Agents need execution that can be verified without exposing everything happening internally.

[Trung Tran @TrungTranlh]

Don't sleep on Wave 3.

It’s the biggest Conservatory wave so far.

If you’ve been waiting for the right moment to push your Season 1 progress, this is it.
@ama_protocol

[Trung Tran @TrungTranlh]

3/3
Amadeus is building around those primitives.

[Trung Tran @TrungTranlh]

@ama_protocol — AGENTIC INFRASTRUCTURE

🧵 1/3
The biggest AI opportunity in finance may not be another chatbot.
It may be infrastructure for agents that actually operate.

[Trung Tran @TrungTranlh]

2/3
Execution, privacy, verification and interoperability all matter.

[Trung Tran @TrungTranlh]

@ama_protocol Meet AMA Agents. 🤖

Agents that can:

→ Monitor markets
→ Analyze data
→ Execute strategies
→ Manage workflows
→ React to events
Not just AI that talks.
AI that acts.

[Trung Tran @TrungTranlh]

3/3
Security and useful computation can reinforce each other.
That's an interesting direction for L1 infrastructure.

[Trung Tran @TrungTranlh]

@ama_protocol — USEFUL COMPUTE
🧵 1/3
Blockchain compute doesn't have to be wasted on meaningless work.
What if network computation could contribute to AI?

[Trung Tran @TrungTranlh]

2/3
Amadeus explores Useful Proof of Work for AI training and inference workloads.

[Trung Tran @TrungTranlh]

What if your trading edge could be used without being revealed?

Amadeus DeFi Agents run strategies inside sealed environments.

Users see the outputs.
The underlying logic stays private.
That changes how strategies can be monetized.

@ama_protocol

[Trung Tran @TrungTranlh]

5/5
Applications are what users see.
Infrastructure is what makes autonomous finance possible.

@ama_protocol

[Trung Tran @TrungTranlh]

Infrastructure Before Applications @ama_protocol
🧵 1/5
Everyone wants to build AI financial agents.
Few talk about the infrastructure underneath them.

[Trung Tran @TrungTranlh]

4/5
Add interoperability and useful AI compute through uPoW.
Now the stack starts looking purpose-built for autonomous finance.