Abstract

Cardano is celebrated for its research-driven design and robust security but faces a critical challenge: translating its technical excellence into a seamless user experience. Fuku AI proposes a bold solution—an on-chain AI agent paired with robust backend infrastructure—to make interacting with Cardano as intuitive as having a conversation. This proposal details our plan to build an intelligent agent that automates Cardano DeFi tasks, which will work in parallel with a powerful data infrastructure we will create, which, together, will make Cardano more accessible for both users and developers. By leveraging Cardano’s native capabilities and emerging AI technologies, we aim to transform Cardano into the most user-friendly blockchain platform.

Achievements on Solana

Over the past six months, a makeshift team of 15 developers has successfully launched Fuku’s AI-driven services on Solana. With over 100,000 lines of code, hundreds of hours of developer work, and a fully self-funded operation, we’ve built a robust platform offering advanced functionalities across DeFi automation, analytics, and portfolio management.

Key highlights include:

• Natural Language Token Swaps: Simplified token swaps and concentrated liquidity management through conversational interfaces integrated with Raydium.
• Automation & Strategy Bots: AI-driven automated trading bots (AutoFuku, HunterFuku) enabling proactive , fully automated onchain investment strategies based on real-time conditions and data analytics.
• Advanced Analytics and Security Checks: Instant access to detailed token analytics, security flags, and real-time charting via Dexscreener integration.
• Portfolio and Wallet Management: Comprehensive wallet management including profit-and-loss analytics, historical funding data, portfolio tracking, and simplified token transfers.
• Automated Trading Tools: Advanced trading utilities such as volume bots, Jupiter limit orders, and AI-driven automated leveraged trading strategies on Hyperliquid.
• Integration with External Data Sources: Real-time integration with Twitter, Telegram, and Pumpfun for instant notifications and automated trade execution based on social signals.

Our backend infrastructure includes powerful Rust and Python services for seamless integration of AI logic, secure key management, and real-time blockchain event streaming. Key components include secure JWT-protected Telegram bots, comprehensive wallet analytics, automated trade execution, and cloud-hosted PostgreSQL databases.

These achievements illustrate Fuku’s proven ability to deliver AI-driven innovation at scale—capabilities we now aim to bring to Cardano.

Vision and Solution Overview

Fuku’s vision is an AI-driven personal DeFi assistant that simplifies interactions with Cardano. Users can simply ask the agent, “Swap 100 ADA for a stablecoin and stake the rest,” and the agent handles the complexities seamlessly. The solution includes:

• AI Agent Layer: Natural language interface executing tasks and educating users in the process.
• Developer Data Infrastructure: Backend services indexing blockchain data, caching, and interfacing with nodes to ensure swift, reliable execution.

By pairing these elements, Fuku delivers a superior user experience, combining user-friendly interactions with robust, secure performance.

AI Agent Architecture

Fuku AI’s robust architecture will consist of:

• User Interaction Layer: Multi-platform client interfaces (Telegram, Web, Mobile).
• AI Logic Layer: Interprets natural language and formulates structured action plans.
• Orchestration Layer: Coordinates task execution, manages transaction flow and errors.
• Integration Layer: Executes specific blockchain interactions through modules connected to Cardano dApps.
• Data & Knowledge Layer: Real-time blockchain indexing from our in-house data infrastructure, providing instant access to essential data and educational content.
• Security & Key Management: Secure wallet integrations ensuring user control over assets, with stringent security policies for transaction execution.

Developer Infrastructure

Our comprehensive infrastructure includes:

• Node Cluster & RPC Layer: Reliable global Cardano node clusters for low-latency blockchain access.
• Custom Indexer: Processes blockchain data into structured formats for rapid queries.
• Caching Layer: Ensures instant responsiveness and scalable performance.
• Developer APIs: Public-facing APIs to support the broader Cardano ecosystem.
• Global Deployment & Scaling: Multi-region, high-availability infrastructure to maintain 99.9% uptime.

This infrastructure not only supports the AI agent but significantly strengthens Cardano’s overall development ecosystem.

Educational Layer

If “ProfessorFuku” is toggled on by the user, Fuku AI will integrate an educational layer directly into future user interactions, enhancing accessibility by turning complex blockchain tasks into learning experiences. The AI agent will proactively provide brief, context-aware explanations during transactions. For instance, when initiating liquidity provision, users might receive a quick explanation about impermanent loss, with the option to ask follow-up questions.

Features of the educational layer include:

• Contextual Tips: Short, timely explanations during user interactions.
• Interactive Explain Mode: Detailed step-by-step guidance for complex tasks, such as staking ADA.
• Personalized Learning: Tailored explanations based on user expertise, ensuring relevance without overwhelming users.
• Standalone Educational Content: Users can directly request structured tutorials on topics like “Getting Started with Cardano DeFi.”

The educational layer will leverage official Cardano documentation and community resources, simplifying complex topics like eUTXO, slippage, and staking. This integrated learning approach aims to transform user hesitation into confidence, significantly enhancing user retention and engagement within the Cardano ecosystem.

Long-Term Impact & Growth Strategy

Fuku AI aims for sustained, transformative impacts:

• Increased User Adoption: Simplified onboarding and reduced friction will significantly grow active DeFi participation.
• Activation of Dormant Capital: Turning passive staking into active DeFi usage, boosting liquidity and ecosystem efficiency.
• Continuous Integration: Quickly adopting new Cardano innovations (e.g., Hydra scaling, governance participation).
• Community Collaboration: Actively engaging stakeholders for continual feedback and ecosystem-driven enhancements.
• Sustainable Monetization: Establishing revenue streams through small fees and subscription models for ongoing improvements. Charging reasonable rates for use of our Public-Facing APIs to fund continuing integrations and ecosystem growth.
• Enhancing Reputation: Positioning Cardano as a leading blockchain by showcasing practical, user-centric AI innovation.

Fuku AI envisions Cardano becoming synonymous with effortless blockchain interaction, significantly expanding its user base and ecosystem vibrancy through reliable, AI-driven automation.

Despite Cardano’s technological strength, UX friction remains a significant barrier. New users encounter steep onboarding hurdles—complex wallet setups, understanding UTXOs, selecting stake pools, and navigating DeFi concepts like impermanent loss. Current wallets and dApps are powerful but intimidating, deterring mainstream adoption.

This friction directly translates into low DeFi activity, with most ADA (over 72%) passively staked rather than actively utilized. Early concurrency issues on Cardano DEXs such as SundaeSwap, slow block confirmation, and multi-step transactions have left lasting impacts on user confidence and activity. Simply put, Cardano’s complexity prevents widespread engagement, leaving vast potential untapped. Fuku AI addresses these issues directly, aiming to simplify and automate interactions, thereby significantly boosting user engagement. It further seeks to simplify the developer experience on Cardano, by providing accessible, easily readable blockchain data to any project that requires it.

Fuku AI’s on-chain agent and underpinning infrastructure promise to introduce lasting benefits for the Cardano ecosystem and treasury. The following points outline the anticipated impact and strategies for long-term, sustainable growth:

1. Higher User Adoption and Retention
   Fuku AI converts the multi-step Cardano onboarding gauntlet (exchange on-ramp → wallet choice → UTXO math → dApp UI learning) into a single conversational flow that finishes in < 30 seconds. If this behavior shift captures even 1 % of inactive Cardano wallets, the tri-monthly count of active DeFi wallets expands from ≈ 40 k today to ≥ 90 k within twelve months—a 125 % jump. Explain-Mode and in-chat safety prompts then cut transaction-error rates, and augment user confidence and comprehension, which will drive measurably higher cohort retention and repeat volume.

2. Activation of Dormant Capital
   Roughly 72 % of all ADA sits passively staked. Fuku’s one-tap “move rewards to yield” templates, real-time APR nudges, and Professor-Fuku micro-lessons remove the cognitive barriers that keep holders out of DeFi. Shifting even 0.5 % of that idle stake (≈ 122 M ADA) into LPs or lending vaults would inject > $55 M in fresh TVL, kick-starting liquidity flywheels for DEXs and generating thousands of extra fee-bearing contract calls each day—all of which feed the treasury.

3. Continuous Integration of Cardano Innovations
   The project’s modular orchestration layer and “super-node” indexer are explicitly architected for plug-and-play upgrades. As Hydra heads, Midnight side-chains, or future CIP-standards go live, new transaction schemas are registered once in the SDK and become instantly accessible via natural-language commands (e.g., “swap via Hydra for instant finality”). This forward-compatibility guarantees that treasury-boosting volume gains persist as Cardano’s throughput and feature-set evolve.

4. Collaboration and Community Involvement
   An MIT-licensed SDK lets any dApp publish functions (openVault, mintNFT, etc.) that the Fuku agent can auto-discover and call, sending additional transactions—and users—to partner protocols without them writing front-end code. Stake-pool operators can embed a “Delegate via Fuku” deep-link that signs a delegation certificate in one click, after which Fuku highlights those pools in its recommended list. Semi-annual hackathons and a free tier of the indexer/API service create a low-friction pathway for builders, compounding the ecosystem impact of every treasury-funded line of code.

5. Public-Facing API Access for Developers
   Running a reliable DB-Sync node cluster plus hot caches can cost a new Cardano project $20 k – $40 k per year. Fuku’s GraphQL/REST endpoints eliminate that hurdle: hobby projects query for free, while production-scale dashboards and analytics sites pay a subscription. The service targets ≥ 25 external dApps/SPO tools in its first year if achieved, those paid tiers alone will spin off a treasury share (5 – 10 % of gross API revenues) that grows in lock-step with overall network usage.

6. Direct Treasury Give-Back Commitment
   The core aim of Fuku AI is to spark materially higher on-chain activity, thereby growing the pool of transaction fees that flow into the Cardano treasury. In addition to this indirect uplift, we formally commit that once our treasury-backed business reaches break-even we will donate a fixed share of our net revenue to the Cardano treasury every quarter. Revenue sources covered by this pledge include (i) paid tiers of our public indexer/API service, (ii) bridge-routing fees on cross-chain transfers executed through the Fuku agent, and (iii) white-label licensing of the AI co-pilot to dApps and wallets. This dual pathway—organic fee generation plus a standing revenue-share—ensures the treasury benefits proportionally as Fuku’s user base and infrastructure revenues scale.

7. Transaction-Fee Flywheel for Treasury Growth
   Fuku’s conversational interface aims to more than double daily DeFi-related transactions on Cardano—from roughly 60 k today to ≥ 120 k within twelve months—by bundling multi-step swaps, LP adds, and staking into single AI-driven flows. Even at Cardano’s modest average fee per transaction, that step-change in volume would translate into a significant, recurring boost to the protocol-level fees that seed the treasury. Moreover, by activating just 0.5 % of the passively staked ADA (~122 M ADA) and funnelling it into yield-producing DeFi positions, Fuku drives more contract calls, reinforcing the fee flywheel.

8. Cross-Chain Liquidity Bridges that Import Fresh Fee Flow
   Cardano’s treasury today captures fees almost exclusively from ADA-native wallets. Fuku bakes first-party Solana ↔ Cardano and Base ↔ Cardano bridge adapters directly into the chat experience, letting users move stablecoins or wrapped assets onto Cardano and deploy them into DeFi in under a minute. Each bridged transfer and subsequent DeFi action is a brand-new transaction the Cardano ledger would not otherwise see, widening the fee base. Even a conservative $20 M influx of cross-chain liquidity (~ > 10 % of current Cardano DeFi TVL) would generate thousands of extra daily transactions—and their attendant treasury share—at virtually no marginal cost to the network.

1st Milestone (40%) - On approval to fund development on Cardano.

2nd Milestone (30%) - When we have basic functionality available for Cardano users. Allow users to swap, trade, and use a number of the above mentioned features with our bot.

3rd Milestone (30%) - When all functionality is completed, and the data source provided is available for paid use by other Cardano developers and projects. Portions of non-proprietary code will be made open source for community audit and forking.

• Three Senior Full-Stack Developers — $300 k / 600 k ADA:
  Market rate. In the US, the median blockchain-developer salary is ≈ $127 k – $160 k; senior full-stack roles routinely top $180 k. We budgeted $100 k each—~35 % below the national senior average—because (i) remote-first hiring lets us tap lower-cost regions, and (ii) part of their compensation is paid in vested ADA. These engineers lead Cardano-specific Plutus integration, perform security-critical code reviews, and mentor six junior staff.

• Six Junior Full-Stack Developers — $360 k / 720 k ADA:
  Entry-level full-stack devs in the US average $75 k – $90 k. We use $60 k to reflect a mix of LATAM/EEA remote talent plus equity incentives. Their mandate: UI/UX refinement, SDK maintenance, API endpoints, and test coverage.

• AI Agent Development & LLM Ops — $150 k / 300 k ADA:
  Covers (a) GPT-4o usage (~15 M prompts × $0.01 = $150 k) for prod inference & fine-tuning; (b) vector-DB hosting for prompt retrieval; and (c) eval tooling (Promptfoo, RAG pipelines). Comparable AI co-pilot stacks (e.g., Solana ChatGPT plug-in) report similar six-figure annual token spends.

• Cardano Backend & Smart-Contracts (external) — $200 k / 400 k ADA:
  Two fixed-scope contracts with specialist vendors for Plutus V2 on-chain wallets + Hydra-ready batcher. Typical Plutus outsourcing quotes run $80 k – $120 k per contract. Budget allows one security-critical refactor cycle and formal verification of invariants.

• Indexing & Data Infrastructure — $150 k / 300 k ADA:
  Annual cloud run-rate for a “super-node”:
  • 6 × m6i-4xlarge indexers ($600/mo each)
  • 3 × read-replica Postgres clusters ($1 k/mo)
  • Global Koios relays & load balancers ($2.5 k/mo)
  • 100 TB/yr data egress @ $0.05/GB → $5 k/mo
  Total ≈ $12.5 k/mo → $150 k/yr. Infrastructure spec matches the Deep-Dive’s multi-region design.

• Cross-Chain Interoperability — $200 k / 400 k ADA:
  Build & audit bridge adapters (Cardano ↔ Solana, Base) + seed liquidity to keep the bridge usable at launch. Remaining funds pay dev days for chain-specific RPC modules. (Liquidity counts as recoverable working capital, but we conservatively list it as expense.)

• Client Development (Telegram / Web / Mobile) — $150 k / 300 k ADA:
  Telegram & Web dashboards are live on Solana; porting to Cardano plus a React-Native prototype needs ~2 devs @ $75 k each, plus design/QA sprints.

• Security Audits & Pen-Testing — $100 k / 200 k ADA:
  Two full Plutus audits at $35 k each (Hacken/Certik quote range: $30 k – $60 k per mid-size DeFi contract) and a $30 k gray-box pen-test of the API & Telegram stack. Hacken’s public methodology notes scope-based pricing in this band.

• Cloud Infrastructure & DevOps — $180 k / 360 k ADA:
  24/7 SRE cover (one senior @ $110 k + on-call pool) and CI/CD, observability, backups. Includes HA Kubernetes clusters for micro-services and an incident budget (PagerDuty, Sentry, Statuspage).

• Educational AI & Onboarding — $100 k / 200 k ADA:
  Creation of 200+ bite-sized “Professor Fuku” lessons, voice-over scripts, and localisation (ES/PT/JP). Cost model: 1 curriculum PM ($70 k) + multilingual copywriters/voice actors ($30 k). Deep-Dive details the Explain-Mode feature.

• Marketing & Community Outreach — $120 k / 240 k ADA:
  • $60 k for four joint AMAs & tutorial videos with top Cardano influencers (avg $15 k per campaign).
  • $30 k for hackathon prize pool & dev-grants to spur plug-in SDK adoption.
  • $30 k for in-person booths at Rare Evo + Consensus. Comparable Catalyst-funded global-event budgets run 1 M for 25 events—our spend is modest in contrast.

• Legal & Compliance — $70 k / 140 k ADA:
  Outside-counsel retainer for US/UK entity structuring ($40 k), privacy-policy & ToS drafting ($10 k), and quarterly regulatory review ($20 k). Benchmarked to rates from DLx Law (NY) for similar DeFi tooling engagements.

• Reserve & Contingency — $20 k / 40 k ADA:
  1 % of total budget. Covers unforeseen node-ops spikes or extra community incentives (e.g., liquidity-mining boosts) without pausing the roadmap. Industry best-practice is 5 – 10 %; we keep it lean to signal cost discipline.

The team already has 7 core developers, 5 community managers, and one CEO/Leader.

We estimate the full project described in the proposal could be completed as soon as six months; this proposal is for 12-months of runway. But we will continue to build after the above has been completed, to make it as robust as possible.

• Punk 2070 (Project Lead): DeFi enthusiast with over 8 years of experience in the cryptocurrency arena. Previous experience on multiple crypto projects that reached market valuations of over $10,000,000. Advised one project which reached a $100 m valuation. Registered securities attorney in New York City.

• Andre (Lead Developer: UI/UX, Backend, Project Management): Quantitative developer with five years of experience. HFT and arbitrage strategy development at an internal market-making desk. Previously a blockchain engineer at J.P. Morgan, experimenting with DeFi strategies.

• Kerry (Senior Developer: UI/UX and Backend): Telecommunications engineer with 7 years of experience as a full-stack software developer. For the past 4 years, has specialized in Web3 software development, primarily on EVM-compatible chains and Solana. Led the development of high-complexity, utility-driven projects such as DEXs, gambling platforms, privacy solutions, and Web3-oriented AI-integrated systems.

• Denis (Senior Developer: UI/UX & Backend): Full-stack software engineer with a degree in computer science and 3 years of hands-on experience in Web3 development. Built and shipped a wide range of projects across EVM-compatible chains and Solana—including NFT marketplaces, DeFi platforms, staking protocols, token launch tools, and dApps with wallet integrations. Stack spans Solidity, Anchor, React, Next.js, Node.js, and PostgreSQL, with a strong focus on clean architecture, scalability, and security. Recently exploring cross-chain infrastructure and integrating AI into decentralized systems to push the boundaries of what’s possible in Web3.

• Team Droid (Frontend & LLM Engineering): Team blending deep expertise in data science, AI/ML, and Web3, led by a seasoned consultant with over a decade's experience advising Fortune 500 firms. Engineering lead is a full-stack developer with hands-on experience building cutting-edge AI products, including contributions to Microsoft Co-pilot and multi-modal LLM research. Combines enterprise-grade strategy with frontier tech execution.

• Mattimouse (Junior Developer; Backend & Cross-Chain Interoperability): Self-taught full-stack developer with 5 years’ experience, predominantly Web3, and more recently at the crossroads of blockchain and AI. Involved in a wide variety of projects ranging from OHM clones to Web3 gaming on multiple EVM-compatible chains as well as Solana. Agentic implementations have taken up the majority of projects over the last 6 months.

• Sailesh / DeSoft Technology (Non-core Contract Development; LLM & Automation): DeSoft Technology is a specialized software firm focused on fintech innovations, with deep expertise in building independent AI models and agentic AI systems. Developed advanced solutions across DEXs, CEXs, and platforms like MT5 and Interactive Brokers for forex and commodities trading, while also collaborating with leading stock brokers worldwide.

• HelloMoon (Non-core Contract Development; LLM Integrations): Blockchain infrastructure provider with deep expertise in supporting complex AI integrations across chains to allow for agentic coordination.

• Lazy Catalyst (LLM Quality Control & Community Management): Software engineer with a computer-science and mathematics degree and a specialization in ML/deep learning. Five years of industry experience doing backend infrastructure at Google.

• Mark (Junior Developer): Final-year student passionate about blockchain and AI integration, with several months of hands-on experience developing in Rust and Python. Currently building expertise in smart contracts and decentralized systems, while exploring AI applications for Web3. Focused on EVM-compatible chains and prototyping AI-enhanced dApps, with a strong foundation in Linux and networking from Cisco certifications.

• Kempa (Marketing / Social Media / Community Management): Marketing lead with a passion for Web3. Four years’ experience managing communities and coordinating marketing campaigns across multiple projects.

Cardano Node Servers (Relay/API), Indexing Servers (DB Sync), Database Instances (PostgreSQL), Caching Layer (Redis/Memcache), API/GraphQL Servers, Load Balancers, Chat GPT, Twitter Premium API, Telegram API.

Because the project will be faced with ongoing costs incurred by its usage of third party APIs (Twitter premium, ChatGPT, server costs) usage of our tools will be paramount to our sustainability. These ongoing costs will be covered by the transaction fees taken by the agent, and the costs we will charge to other projects rutilizing our API to fetch the blockchain data we will index and expose.