Portaldot Hackathon · 2026 · AI Track

Verity
Onchain
Intelligence

Blockchain data is locked behind developer tooling. Verity fixes that. Ask any question in plain English — Gemma AI routes it to 20+ live chain queries and returns structured data instantly, with every query logged on-chain.

Live Query Preview
top 20 POT holders
get_top_holders(max_results=20)
23 ms ⚡ cached
Address Free POT Total POT
5GrwvaEF…GKutQY 847,293.00 847,293.00
5FHneW46…M694ty 421,847.50 421,847.50
5FLSigC9…jh5T6J 89,234.00 89,234.00
5DAan…SLZHT 12,400.75 12,400.75
✦ AI Analysis
Alice controls 847K POT — 63% of circulating supply. High concentration: top 3 accounts hold 98% of all tokens.
20+
Query Types
AI
Powered Routing
100%
Live Chain Data
0
SQL / Code Required
01. The Problem

"Blockchain data is locked behind developer tooling. Non-technical users have no access."

— The barrier we eliminate

To answer a simple question like "who holds the most tokens?" requires knowing the SDK, pallet names, data structures, and how to write Python against an RPC node. Investors, auditors, compliance teams, and business analysts are locked out entirely.

Existing block explorers show raw data but don't answer questions. They require you to already know what you're looking for. The barrier to reading blockchain state shouldn't be knowing SDK method names.

This is what that problem looks like — solved.

02. How It Works

Three steps.
No code.

01
Ask in Plain English

Type any question about the Portaldot blockchain — "show me the top token holders", "are there stuck multisig proposals?", "how much POT moved in the last 100 blocks?" No special syntax required.

02
Gemma AI Routes

Google's Gemma 27B maps your question to the correct one of 20+ chain query functions. It extracts parameters like block ranges and addresses automatically — you don't specify them unless you want to.

03
Chain Responds

The Python SDK queries live Portaldot state over WebSocket. Results are cached, charted, and returned in milliseconds. Every query also logs itself on-chain as a signed extrinsic, paying real POT gas.

03. Capabilities

What you can
query right now.

Token Analytics

Top holders, total supply, token concentration, wealth distribution, transfer volume, and largest individual transfers.

Transfer Intelligence

Full transfer history, active address counts, transfer volume aggregates, and whale transaction detection across any block range.

Multisig Monitoring

Detect stale pending proposals, rank multisigs by locked POT, track the full lifecycle from creation to execution or cancellation.

Network Stats

Extrinsics per block, average block time, fee statistics, live block feed, and per-block deep dives showing every extrinsic.

AI-Generated Insights

After every query, ask Gemma to interpret the data. Get 2–3 sentences of sharp analytical commentary — patterns, anomalies, risks.

On-Chain Audit Trail

Every query logs a System.remark_with_event extrinsic signed by Alice, paying POT gas. All analyst activity is permanently on-chain.

04. Architecture

Built on
real infrastructure.

Browser
  ↓  POST /query
FastAPI Backend
  ↓  parse_query()    → Gemma 27B (OpenRouter)
  ↓  sdk_runner.*()   → substrate-interface
  ↓  log_query_onchain()
Portaldot Node  (WSL · ws://127.0.0.1:9944)
  Substrate Layer-0 · V13 metadata
  POT token · 14 decimals · --dev --alice
  • Portaldot Substrate-based Layer-0, native POT gas token
  • Gemma 27B Google's LLM via OpenRouter for NL routing + insights
  • substrate-interface Python SDK for live WebSocket chain queries
  • FastAPI Backend with in-memory caching and 20 query endpoints
  • Chart.js Line, bar, and doughnut charts for visual data analysis
Ready to query the chain?

Start in
30 seconds.

No configuration needed. The dashboard connects to the live Portaldot dev node automatically. Type a question and get answers.

Launch Dashboard