kwikwurk

An AI-native, multi-agent Web3 job platform that uses on-chain history as a cryptographic resume and deploys automated scrapers to find real-time freelancing gigs from new token launches and X post

Comment

## Inspiration

The Web3 ecosystem operates at extreme, borderline-chaotic speeds. Every single day, hundreds of early-stage startups and token initiatives deploy code directly to public ledgers or quietly announce their "stealth launch" on X (Twitter). They desperately need developer talent, auditors, designers, and community builders right now.

Yet, traditional job platforms are slow, flat, and fundamentally broken for this ecosystem. Talent waste hours formatting static PDF resumes that fail to accurately reflect their real engineering competencies. Meanwhile, founders spend tens of thousands of dollars screening unverified applications.

We built KwikWurk to eliminate this frictional gap. We asked ourselves: What if a freelancer’s public wallet address became their ultimate, self-verifying, cryptographically proven resume? And what if an AI could monitor the blockchain live to alert freelancers about paying gigs before the founders even have time to draft a formal job ad? KwikWurk turns this concept into a live, high-velocity reality.

## What it does

KwikWurk is an autonomous, AI-native Web3 talent engine designed to match on-chain expertise with live protocol needs in real-time.

  • The Cryptographic Resume Parser: When a freelancer connects their public wallet address, KwikWurk doesn't just read an address; our AI backend deep-scans their public ledger history. It aggregates smart contract deployments, DeFi protocol interactions, DAO governance voting records, and transaction frequency to auto-generate a verifiable skill map.
  • The Alpha Scout Bot Feed: Our automated tracking pipeline constantly monitors live blockchain nodes and targeted X streams. The moment a new token contract is deployed or a stealth team signals a launch, KwikWurk lists the project instantly on our pre-login homepage dashboard, flagging an immediate opportunity.
  • Premium Multi-Channel Autopilot: Users on our Premium Tier receive instant, ultra-low-latency matching notifications via Telegram, SMS, or Push notifications the second an on-chain deployment aligns with their verified skills. Furthermore, the system leverages Gemini to auto-generate context-aware, completely tailored cover letters customized to that target protocol's exact architecture.

## How we built it

To fulfill the strict engineering criteria of an AI-native operational platform, KwikWurk leverages an integrated, highly scalable Google Cloud and AI stack:

  • Google AI Studio & Gemini API (Gemini 3.5 Flash): This serves as the core intelligence layer. We use it to calculate vector embeddings that map raw on-chain transaction data to our database of primary Web3 roles. It is also the generative engine behind our real-time, context-aware cover letter builder.
  • Google Cloud Run: Hosts our responsive frontend multi-page application built utilizing Next.js (App Router), TypeScript, Tailwind CSS, and Web3 injection libraries like AppKit/RainbowKit. Cloud Run allows the app to scale instantly to handle fluctuating user loads.
  • Google Cloud Pub/Sub & Cloud Scheduler: Manages our continuous, decoupled cron-triggers that drive our automated scraping agents, ensuring our Alpha Tracker remains completely live without requiring manual human oversight.
  • Supabase / PostgreSQL: Acts as our secure database infrastructure to handle encrypted user profiles, application historical states, and transactional ledger states.

## Challenges we ran into

Integrating raw blockchain telemetry with real-time semantic analysis presented a steep technical learning curve.

  1. On-Chain Data Noise: Raw transaction histories are incredibly noisy. A developer wallet might contain thousands of generic swap functions, spam tokens, and minor interactions. Fine-tuning our Gemini prompt strategies to effectively separate meaningful behavioral footprints (like deploying a custom ERC-20 or voting in a major DAO) from background noise required rigorous testing.
  2. Scraper Latency: Building an "Alpha Tracker" means milliseconds matter. X API limits and block confirmation times initially caused bottlenecks. We solved this by implementing parallelized data ingestion pipelines via Google Cloud Pub/Sub, separating the raw tracking logic from the Gemini processing step.
  3. Handling Unauthenticated UX: We had to carefully architect a high-conversion landing page that demonstrates real value before a user links their wallet. Designing interactive web states that cleanly blend real-time public feeds with locked, premium matching layers required highly tactical frontend state management.

## Accomplishments that we're proud of

  • True Zero-Input Profile Building: We built a platform where an applicant can securely sign in with an empty profile, click "Connect Wallet," and watch an advanced, multi-page professional identity populate automatically within 10 seconds based entirely on their cryptographic history.
  • End-to-End Autopilot Pipeline: Our background agent system functions seamlessly on its own. It uncovers new contract deployments on Base and Solana, passes the project data to Gemini to extract intent, checks our database for matching profiles, and fires off real-time Telegram alerts entirely without our intervention.
  • Operational Profitability Inside the Window: We designed a sustainable business model from day one, establishing explicit pricing structures that generate verifiable commercial traction directly inside the hackathon testing timeframe.

## What we learned

  • The Power of Residual Context: Building with the latest Gemini models taught us that LLMs are remarkably adept at identifying developer intent when provided with structured transaction schemas. It can recognize the structural difference between an amateur hobbyist project and a complex production dApp deployment.
  • UX Trumps Hype: In Web3, users are weary of empty promises. We learned that the most effective way to drive a wallet login is to show functional, real-time public utility up front, utilizing features like our unauthenticated Live Alpha Tracker to immediately prove the platform's mechanical value.

## What's next for kwikwurk

The hackathon submission is just our initial baseline. As we scale KwikWurk into a permanent Web3 venture infrastructure, our immediate operational milestones include:

  • Multi-Chain Indexing Scaling: Expanding our background tracker beyond EVM networks and Solana to natively support burgeoning Move-based architectures (Sui/Aptos) and specialized Layer 2 environments.
  • On-Chain Escrow & Milestone Smart Contracts: Integrating native cryptographic payroll solutions, allowing clients to fund a job posting directly into an escrow contract that safely auto-releases funds via AI-verified milestone completion.
  • Deepening the Developer Community Ecosystem: Partnering directly with global Web3 developer DAOs and layer-1 foundations to integrate KwikWurk as their primary, native talent scouting system.

Built With

Share this project:

Updates