From raw on-chain data to real-time trading intelligence.

A multi-agent crypto analytics assistant that sits on live wallet and transaction streams and answers questions grounded in exactly what the trader has selected.

Discuss a similar build All case studies

Client: Trenches.ai
Industry: Crypto / DeFi / Trading Intelligence
Delivery: Web platform and multi-agent AI system
Capability: AI Agents & Automation

The challenge

What Trenches.ai was up against.

Active traders juggle dashboards, raw transaction logs and generic chatbots that know nothing about their actual positions.

Real-time transaction and wallet changes
Manual interpretation of raw logs
No link between wallet activity and market context
Generic answers without transactional grounding

The solution

What we built.

Skugal engineered cooperating agents on top of live data, so the UI, the data and AI reasoning stay in one tight loop.

Live agent activation & sync

The agent pulls wallet and API data and the UI updates itself as on-chain state changes.

Context-aware chat

An inline assistant for portfolio, P&L, token details, swaps and trends, with follow-ups that keep context.

Transaction-scoped questions

Select a transaction and every answer is grounded in it until the context changes.

Portfolio & trend intelligence

Portfolio breakdowns, P&L and trending tokens fetched from on-chain providers and structured for the UI.

How it works

From input to outcome.

UI interaction & selection
Query classification agent
Context assembly
Vector retrieval (Pinecone)
On-chain API calls
Response structuring agent

Business impact

What changed.

Trenches.ai became a living crypto intelligence interface that moves traders from charts to agent-assisted decisions.

Technology

Prompt-orchestrated LLM agents
Pinecone
Solana & wallet APIs
Webhooks
Reactive real-time UI

Data that updates itself
AI that understands user intent and selection
Questions asked inside the trading workflow
Insights tied directly to real transactions

It doesn’t just show what happened on-chain. It lets traders ask why.

More case studies

Related work.

Token MetricsTurning crypto market noise into personalised investment signalsExplore

Plato TechnologiesAn AI-driven crypto platform for price tracking, portfolio insight and trendsExplore

LendlyAutomating complex lending operations with agent-driven workflowsExplore

Next step

Have a similar challenge?

Tell us about the workflow or product. We’ll come back with an architecture, a delivery plan and where AI agents fit.

Contact our team Talk to our AI voice agent

Or email ceo@skugal.com