IdeaTON: How To Hack

From TON Wiki (En)
Aleksei Solovyov, organizer and host of IdeaTON

«IdeaTON: How to Hack» is a closed event organized by TON Society CIS and DeDust.io for project creators and developers in TON. It took place on March 23, 2025, in Saint Petersburg at the «Yasnaya Polyana» coworking space.

The goal was to bring together active members of TON Ecosystem in a live setting to generate ideas and present new projects ready for development and implementation.

Event Format

Anyone could apply via the ONTON bot. Confirmed applicants could claim a commemorative SBT. A total of 55 participants attended. Entry was free of charge.

Tracks and prize pool of the hackathon:

  1. AI Agents (1st place — $750, 2nd — $500, 3rd — $250)
  2. DeFi (1st place — $750, 2nd — $500, 3rd — $250)

The DeFi track was sponsored by the decentralized exchange DeDust.io.

Total prize pool: $3000

The event was organized, hosted, and moderated by Aleksei Solovyov, TON Society CIS Hub Leader.

«IdeaTON is our most successful format by KPI — after such events, we often see participating projects grow further» (Aleksei Solovyov)

IdeaTON2 EN.jpg

Event Structure:

  • Opening remarks by the host
  • Presentations by four speakers on hackathon tracks
  • Lunch break
  • Presentation preparation
  • 3-minute pitches from participants; Jury questions, 5 mins each
  • Announcement of winners

Speaker Presentations

IdeaTON3 EN.png

1) Andrey Shvets, smart contract developer at TON Studio, presented on AI agents in TON and the shift from imperative to declarative interaction with AI. Since an AI agent is autonomous and fully integrated, a user can not only ask «where to invest,» but also set up auto-execution of the strategy.

2) Aleksei Solovyov introduced TON Nest, prepared by Alik, Builder Programs Lead at TON Society.

TON Nest is a three-tier «pre-accelerator» for emerging projects offering support at each level:

  • Level 1 – Pitch Deck
  • Level 2 – MVP
  • Level 3 – Traction

Level 3 unlocks access to grants, tech credits from partners, and marketing support. As of now, there are 700+ projects and ~1000 applications submitted to TON Nest. Apply here

3) Nikolay Novosyolov from Frens Community showcased the potential of multi-agent systems and specific AI models.

Use cases of AI agents include:

IdeaTON4 EN.png
  • DeFi — e.g., SingularityDAO uses AI to manage crypto portfolios
  • Healthcare — e.g., Doc.ai analyzes health data and provides personal insights
  • Personal Assistants — e.g., Autonomous Life Assistant manages budgets via smart contracts
  • Supply & Logistics — e.g., Lightchain Protocol AI forecasts demand and optimizes routes
IdeaTON5 EN.png
  • Gaming & Metaverses

- Virtual AI NPCs with crypto wallets - GameFi Yield Optimizers - NPC Guilds earning crypto collectively

4) Vitaliy Kitaev, CMO of DeDust.io, explained liquidity pool mechanics and the upcoming protocol DeDust X — an Uniswap v4 analog with simplified logic and customizable pools via Hooks.

  • Each pool can be linked to one Hook — a separate smart contract with custom logic
  • One Hook can be used for multiple pools
  • Hook is attached at pool creation

Hackathon Winning Projects

There were 8 project submissions. Some teams formed on-site. The winners were:

Ads AI by @v_litvintsev

IdeaTON6 EN.png

Problem: Telegram ad campaign setup is a manual, time-consuming task for marketers.

Solution: An AI agent replaces Telegram marketers, reducing costs by ~90% with:

  • Full autonomy
  • 24/7 campaign management and market monitoring
  • Memory of past campaign performance

Taurus by @donlusofon

IdeaTON7 EN.png

Problem: Many Telegram users don't use crypto but would love passive income.

Solution: Margin farming with Telegram Stars

User flow: The user deposits Stars, selects a lock period, and receives interest + deposit in $USDT.

Taurus revenue:

  • Lending interest
  • Early withdrawal penalties

No direct crypto interaction is needed — DeFi access is simplified via Telegram Stars.

Decentrify by @iaroslav444

Problems: Portfolio optimization is complex; building AI agents is hard for non-tech users.

Solution: A no-code DeFi platform for AI agent creation and execution.

Features:

  • Real-time strategy adjustment
  • Multi-agent collaboration

CITY HEMP @cityhemp

Problem: Lack of transparency in plant cultivation processes

Solution: A high-tech agro-crypto company using AI to optimize growth

AI assists with:

  • Growth data analysis
  • Process recommendations
  • Ensuring only legal crops are grown

End result: Users monitor their farms via a Telegram mini app while AI handles the operations.

Links:

  1. TON Society Website
  2. TON Society CIS Telegram Chat (RU)