Author: Gandalf, Techub News
Introduction
In the on-chain market, it is not hard to see price discrepancies across different trading venues; the real challenge lies in assessing whether these price differences can be executed under real conditions, and if there will still be verifiable results after accounting for slippage, fees, gas, capital costs, path, and time costs.
A computer science graduate, who will graduate in 2025, has previously participated in DApp development and is now working as an independent developer and agent development engineer focused on the intersection of AI and Web3. He is currently building an Arbitrage Dashboard — a visual workspace that integrates market observation, on-chain trade research, and arbitrage opportunity analysis. The project aims to consolidate exchange pricing, on-chain data, and transaction records into a single research interface, allowing users to not only “see data changes” but also track where the funds come from, what paths they take, how costs are structured, and whether a historical opportunity is reproducible.
From his perspective, AI's role in on-chain research should not be to generate seemingly reasonable analyses but should be built on real data and deterministic calculations, helping researchers to quickly organize complex information, form evidence chains, and eliminate opportunities that have apparent price differences but are practically non-executable or even lead to losses.
Core Message
“Price differences only indicate that two places have different quotes at a given moment; they do not represent a profitable opportunity.”
The focus of the Arbitrage Dashboard is not to execute trades directly, nor to promise the discovery of arbitrage opportunities, but to distinctly differentiate between “market quotes,” “candidate opportunities,” and “verified results.” The project currently operates within the boundaries of read-only research and simulation verification, helping users move from discovering clues to validating conditions rather than packaging a momentary price difference directly into a conclusion.
For him, the greatest value of AI is not just to find opportunities faster, but to answer more quickly why this opportunity exists, what conditions could render it invalid, and whether it is still worth further research after accounting for all costs.
Starting from Real Pain Points
TECHUB NEWS: Please introduce yourself and the Arbitrage Dashboard you are developing.
Small White: I am Small White. Previously, I participated in DApp development in the Web3 field, and now I mainly work on independent development at the intersection of Web3 and AI, as well as being an agent development engineer.
The Arbitrage Dashboard is a workspace for on-chain research focusing on arbitrage. It aims to organize scattered market data and on-chain data into combinable and traceable arbitrage research conclusions, allowing researchers or traders to complete the necessary information acquisition, path organization, and opportunity validation in one workspace.
This is not about having the system place orders directly for users, but rather hopes to reduce the information-switching costs during the research process: when users see a price difference, they will be able to further understand the capital flow, trading paths, cost structure, and executable conditions, rather than just staying at a single number.
TECHUB NEWS: Why do you prioritize AI application development and Web3 on-chain research as your focus exploration directions?
Small White: These two directions align with my interests and technical background. Many data in Web3 naturally exist on-chain; in the past, researching a transaction or a capital path often required manual extraction, comparison, and organization of information from different pages, which was costly.
AI is particularly adept at handling complex, dispersed, and unstructured information, so there is an opportunity to assist researchers in processing work that originally required a lot of manual effort. However, I do not want AI to merely generate a seemingly reasonable analysis but hope that it is based on real data and deterministic calculations: by obtaining sufficient context, it can help users understand why an arbitrage opportunity exists, why it may be non-executable, and ultimately why it may not lead to profits.
Putting Dispersed Data into the Evidence Chain
TECHUB NEWS: The market already has market analysis software, blockchain explorers, and DEX data platforms. Why do we need a new visual workspace?
Small White: This stems from my own research pain points. When verifying an on-chain transaction, one often has to repeatedly switch between blockchain explorers, transaction detail pages, token transfer records, internal call information, and different protocol pages. Existing tools can inform users of the data changes that occurred, but it is difficult to directly answer several key questions on the same page: Where does the capital come from? What trading paths were taken? What is the total cost? Can the profits be verified? Is this a reproducible opportunity?
Constant switching between pages incurs costs and makes the research process fragmented. The Arbitrage Dashboard hopes to organize this information according to the logic of trades and capital into a sequential, traceable research workspace. Visualization is not intended to make the data more complex but to help people quickly see the relationships between paths, costs, and evidence.
TECHUB NEWS: From “discovering price differences” to “confirming real opportunities,” what difficulties still exist?
Small White: If you randomly open two DEXs or one DEX and one CEX, you can often find price differences at a given moment. However, price differences only indicate that two venues have different quotes, and do not mean that users can transact at the expected scale, nor that they will ultimately make a profit.
Further verification of multiple conditions is needed: How much price impact will this transaction bring; do the liquidity pools at both ends have sufficient depth; how are fees, gas, and capital costs calculated; is the trading path feasible; if cross-chain is involved, will the cross-chain time change the price conditions; is the account capacity sufficient; and can the transaction be timely packaged and successfully added to the chain in a competitive environment.
When these conditions and costs are deducted one by one, many apparent opportunities could turn into losses or may not be executable at all. Therefore, what the project aims to solve is not to “discover price differences faster,” but to “systematically verify whether opportunities truly exist.”
Quotes, Clues, and Verification Results
TECHUB NEWS: How can we simply differentiate between market quotes, candidate arbitrage opportunities, and verified results?
Small White: Market quotes are the most basic information: a certain asset shows different prices on different platforms or liquidity pools at a given time. It can be a clue but not a conclusion.
Candidate arbitrage opportunities are a preliminary listing of trading direction, potential paths, and anticipated profits after noticing a price difference. At this stage, hypotheses need to be proposed, such as where to buy, where to sell, which protocols or links to go through, and what the expected scale is.
Verified results must incorporate executable conditions into the calculations and validations: actual executable scale, slippage, liquidity depth, fees, gas, capital costs, cross-chain or settlement time, trading competition, and path risks, among others. Only when all these factors are reviewed can one determine whether a clue still has research value.
These three cannot be conflated. If the quote is directly treated as achievable profit, users may easily overlook the real execution environment; if candidate paths are directly treated as verified results, they may incur risks and costs that have not been accounted for behind seemingly favorable numbers.
TECHUB NEWS: In the Arbitrage Dashboard, what is the ideal research process from discovering clues to completing verification?
Small White: First, users discover potential price differences or anomalous signals from market information such as exchange quotes, funding rates, and DEX liquidity; then, they treat it as a candidate opportunity and outline possible trading directions and paths.
Next, the research focus shifts from “what is the price” to “can it be executed”: checking the pool depth and expected slippage, estimating fees and gas, analyzing whether capital needs to cross-chain, what effects cross-chain time and paths will have, and combining account capacity and competition circumstances to evaluate potential limitations.
If the research object is a historical on-chain transaction, it is also necessary to further trace where the funds came from, through which contracts or protocols they passed, what costs were incurred along each path, and whether the final capital flow matches the results. Only when the paths, costs, capital flows, and final results can correspond to each other can researchers form a more complete evidence chain.
The project currently focuses on research and simulation verification to help users understand and authenticate these conditions without executing real trades.
AI as a Research Assistant, Not a Profit Promise
TECHUB NEWS: As an agent development engineer, what roles do you think AI and agents are best suited to perform in products?
Small White: AI and agents can assist in clue discovery, data organization, path explanation, information comparison, and proposing hypotheses to be verified. Given the large volume of on-chain transaction records, protocol calls, and market data, machines can more quickly help consolidate dispersed information into a single research path and prompt users about which data is still missing, which costs may have been overlooked, and where conflicts exist.
However, the ultimate goal is not to have AI provide an answer that “can definitely be arbitraged,” but to have it assist users in establishing a judgment process based on real data and deterministic calculations. Particularly in an on-chain environment, where data sources are dispersed, timeliness varies, and trading conditions change quickly, AI should help users identify incomplete information and false opportunities rather than packaging uncertainty into overly confident conclusions.
TECHUB NEWS: In on-chain research, is AI's more important value in discovering opportunities or in eliminating false opportunities?
Small White: Both are important, but for researchers, eliminating false opportunities is equally critical and often has more practical value. Many instant price differences can be seen in the market, but many no longer have profit margins after considering slippage, fees, on-chain congestion, time, and competition.
If AI is only responsible for pushing more opportunities without the ability to explain their executable conditions, risks, and costs, users' decision-making burden could increase. A more meaningful workflow is: AI assists in discovering clues, while also helping to propose counter-questions and verification checklists, converting “what seems doable” into clear explanations of “why it can be done or why it cannot be done.”
TECHUB NEWS: Why does the project choose to operate within the boundaries of read-only research and simulation verification, rather than executing real trades directly?
Small White: There are many dynamic conditions between price differences and real execution. Real execution involves risks such as capital, account security, transaction timeliness, slippage, on-chain congestion, and market competition. For a research-oriented tool, clarifying data, calculations, paths, and evidence chains is more important than directly involving users in real trading.
The positioning of read-only research and simulation verification also helps users understand the product's output boundaries: it provides research support and condition verification, rather than profit promises or automatic trading instructions. By prioritizing “research reliability,” the project aims to establish a more cautious foundation between product credibility, transparency, and user risk management.
Conclusion
For Small White, the value of the Arbitrage Dashboard lies not in packaging instant price differences into an automated profit story, but in breaking down an on-chain research clue: Is the quote real, is the path feasible, how does the capital flow, can the costs be verified, and does the trade have executable potential in the real environment?
In the context of the accelerating intersection of AI and Web3, on-chain data does not inherently equal clear conclusions. AI can reduce the costs of manual organization and cross-platform verification, but more importantly, it helps researchers build evidence chains, identify missing conditions, and distinguish market signals from verifiable results.
From “seeing price differences” to “verifying opportunities,” there are not only technical connections but also a meticulous examination of costs, liquidity, time, competition, and risks. For the Arbitrage Dashboard, which is still in the development and verification phase, this emphasis on research boundaries is also the core value it hopes to deliver to users: first clarify the opportunities before discussing whether to take action.
Editor’s Note: This article is based on the interview text, meeting summary, and interview outline of Small White by TECHUB NEWS. The product positioning, functional ideas, data research processes, and risk boundaries related to the Arbitrage Dashboard are all based on the interviewee's statements and provided materials. The project is still in the development and verification phase; relevant content does not constitute recommendations, predictions, or guarantees for any trading strategies, digital assets, protocols, trading platforms, or market trends.
Disclaimer: This article is for informational exchanges, technical discussions, and research methodology explanations only and does not constitute any investment, trading, or arbitrage advice. The prices of digital assets and related products can rise and fall, and on-chain transactions may involve risks such as smart contracts, liquidity, slippage, gas, cross-chain, network congestion, counterparty, and security. Market quotes, historical transaction records, simulation results, and any analyses do not represent future realizable trading results. Readers should not rely solely on the content of this article to make investment or trading decisions and should assess their own goals, financial situations, risk tolerance, and applicable laws, consulting independent professional advice as necessary.
免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。
