LAKUZO
The World 1st Mobile Prediction Market Aggregator
Created on 12th December 2025
•
LAKUZO
The World 1st Mobile Prediction Market Aggregator
The problem LAKUZO solves
Prediction Market is Predictable!
As of late 2025, Prediction market weekly notional volume has surpassed $4 Billion, with platforms like Polymarket, Kalshi, and Opinion Labs settling billions in liquidity every week. The market has found product-market fit.
But the user experience is stuck in 2013. Currently, a trader trying to track a major event (like the US Election or a Fed Rate Cut) has to navigate a fragmented landscape. They keep five tabs open:
- Polymarket for global liquidity.
- Kalshi for US-regulated contracts.
- X (Twitter) for real-time news.
- Dune Analytics for volume data.
- Spreadsheets to manually calculate arbitrage.
There is no "CoinMarketCap" for this sector. There is no single "terminal" where you can see the truth across all chains and platforms in one view. That is the gap Lakuzo fills.
Phase 1 (Live currently):
We are not a new betting market; we are the intelligence layer that sits on top of them.
1. The "CoinMarketCap" Feed
We normalize odds from disparate sources (currently Polymarket & Kalshi) into one standard, real-time feed.
2. "Internal" Arbitrage Discovery
In an efficient market, the price of "Yes" + "No" should always equal $1.00 (or 100%). However, due to liquidity crunches or panic selling, these markets frequently break.
- The Opportunity: We often see markets where Yes ($0.55) + No ($0.40) = $0.95. This is a risk-free 5% yield.
- The Solution: Lakuzo’s engine scans for these specific inefficiencies within each platform and highlights them.
3. Cross-Platform Analysis (The Workspace)
We are in the process on refining the automated the direct pairing of Polymarket vs. Kalshi arbitrage (e.g., Buy Yes on Poly, Buy No on Kalshi), But we built the tool that makes manual discovery possible: The Bundle Workspace.
- Lakuzo allows users to search for "Bitcoin" and drag charts from both Polymarket and Kalshi into a single "Bundle." This creates a custom dashboard where a trader can visually monitor price divergence between platforms in real-time, effectively building their own arbitrage station.
4. The Insight: Why Social Sentiment is Broken
- While building Phase 1, we analyzed existing "Social Sentiment" tools like Hashdive (Smart Score) and identified a critical flaw in how on-chain data is interpreted.
- Existing tools track "Smart Money" by looking at wallet profitability and volume. But on prediction markets, high-frequency traders often execute hundreds of micro-orders on a single "sure thing" position.
- The Distortion: If a whale buys "Yes" 300 times on a 99% odds market, legacy tools count that as "300 Smart Signals." It bloats the data. It tells you who has the most volume, not who has the best foresight.
- The Consequence: Retail users get tricked into following "high win-rate" wallets that are actually just farming easy/safe volume, rather than making insightful predictions.
5. The "One Person, One Vote" Reputation Protocol
- We are introducing a reputation layer that lives entirely on Base where users vote on market outcomes without betting capital. Instead, they stake Reputation Score. You have to vote once to view that specific market Yes and No Reputation Score and Sentiment (it's a good practice to bid here before committing on a real poly/kalshi trade like a demo trading account on real market data)
- The Anti-Whale Mechanism: You can only vote once per market. Whether you have $10 or $10 Million, your vote counts as one signal. This prevents "stat padding" and creates a cleaner signal of human sentiment vs. capital sentiment.
- Scoring: +10 Rep for correct predictions, -10 Rep for incorrect ones. Over time, this filters out the noise and highlights the true top forecasters.
Phase 2: Make Prediction Market More Predictable
Phase 2 shifts Lakuzo from a passive "Viewer" to an active "Protocol." We are building the social and financial incentives to fix the data flaws mentioned above.
1. The Incentive: Turning Gas into Gold
- This is where Base becomes our competitive advantage.
- The Cost: Voting requires an on-chain transaction. On Base, this costs ~$0.0003 (negligible).
- The Pivot: In Phase 2, we will increase this cost slightly (e.g., to $0.001). That extra fraction is a "Protocol Fee."
- The Payoff: 100% of these fees go into a Seasonal Prize Pool (Smart Contract). At the end of the season, the accumulated pot is airdropped to the Top 10 users on the Leaderboard.
- Why this works: We turn spam into yield. Every bot or noisy user that pays to vote is effectively funding the reward for our smartest users.
2. The "Pro" Terminal
Finally, we are bringing professional charting to prediction markets. We are integrating a TradingView-style engine that allows users to draw trendlines, RSI, and MACD on probability charts.
Challenges I ran into
1. The "Zombie Market" Problem
We wanted Lakuzo to be "Market Agnostic," but we quickly realized that simply importing everything made the platform look like a junkyard. Both Polymarket and Kalshi have thousands of "Zombie markets" with $0 liquidity or niche topics that never started.
- The Decision: We built a strict filter. We don't want to show dead data.
- The Logic: On Polymarket, we set a hard floor: the market must have over $5,000 Volume. For Kalshi, since they report in "contracts" rather than USD, we had to do some rough math. Assuming an average contract price of ~$0.50, we set the threshold at 10,000 Contracts to roughly match the $5k liquidity standard. This creates a clean, tradeable feed from Day 1.
2. Designing "Honest" Sentiment
We curate from many prediction market tools and looked at Hashdive’s "Smart Score" for inspiration, but during development, I noticed a flaw in their logic. They lean heavily on on-chain volume.
- The Flaw: If a whale spams 300 orders on a 99% odds market, legacy tools count that as "300 Smart Signals." It rewards Volume Farmers, not Forecasters.
- Our Fix: This forced us to pivot our product design. We built the "One-Person-One-Vote" Reputation Protocol. By decoupling "Money" from "Voice," we ensure that a whale with $1M has the same voting weight as a user with $10. We want to track skill, not wallet size.
3. API Limits vs. Real-Time Needs
We promise Arbitrage discovery, but you can't arbitrage old data. Relying on external APIs meant hitting rate limits if we queried too fast, or serving stale data if we queried too slow.
- The Solution: We built a Hybrid Scraper.
- Background: Our server runs a safe, steady scrape every 5 minutes to build the baseline database without getting banned.
- Foreground: The moment a user clicks a specific market, we trigger an Atomic Live Refresh just for that asset.
- The Result: The user always trades on the exact second's price, but we don't hammer the API for the thousands of markets no one is looking at.
Tracks Applied (1)
