跨學科領域交織的美妙之處，在於能夠讓人開創新思維，和以更多不同角度理解問題的本質。
故事要從我在約翰肯尼迪政府智庫學院（John Kennedy School of Government, Harvard University）學習生涯說起，當時我擔任博士後研究員院士，期間學到了很多有用知識。每當我與同儕討論任何問題，無論是源自商界、政府，還是慈善組織，都會涉及原子能、政治學、工程學、企業重組等範疇，總能找出最優的解決方案。
在某種意義上，STEM的概念目前仍處於初階，正如福爾摩斯（Sherlock Holmes）所言：最佳的描述可能只是「開端」。
讓我們延續第一篇電影《功夫熊貓》（Kung Fu Panda）的故事，兒子問父親：「秘竅是什麼？」事實上，數據智能的秘密成分之一是「逆向工程」（reverse engineering）。
逆向工程定義
逆向工程是一個過程，對我來說，它也是一種心態、思維和演繹推理的方法，透過洞察力剖析產品、系統、原型或以前解決方案中的開發和操作原理，由此我們可以發現差距，進一步改善弱點和以前被忽略、不可操作，甚至不存在的關係。
讓我們重溫這句話，你會發現它給了我們洞察力，就是數據智能的關鍵，也是本文的主旨和重點：數據智能和逆向工程。
長期以來，逆向工程一直應用於連接不同軟件系統的概念驗證：例如Microsoft WindowsSCOPE，它展示所有文件和APIs的完整內容，以突顯各自的競爭優勢，並重組丟失或消失的原始文件，所以它並非什麼新鮮事。
另一個很好的應用個案發生在貝爾實驗室（Bell Laboratories），他們曾在Apple Macintosh SE上的原型機器以外的其他系統上成功使用逆向工程。
數據智能中的逆向工程
對於如此多的數據項目，但迄今在數據智能方面，還沒有嘗試利用逆向工程，我反覆自問：究竟真正問題是什麼？
那些付費為數據解決方案的公司，尤其是那些為數字化轉型項目付費的公司，為何最終走向失敗？
真正答案是這些公司想要做的只是提升附加值（added value），但在轉型項目上是通常很難成功。
我經常發現公司內部有着很多個不同的筒倉/孤島（silos）（有些不為其他人所知）。往往在項目工程後期，需要制定解決方案時，才發現這麼多數據紕漏，最終需要回到原來各自筒倉/孤島的起點。
為了省卻大家的時間，我寫了這個題目，這是部分原因，而更重要的是，我也制定了數據智能中的逆向工程流程，希望與公眾分享。
數據智能變得簡單
- 存在的問題：大多數數據項目顧問從左到右工作，那麼永遠不會達到「價值終點」。
- 解決方案：秘訣是從右向左工作，使終點和起點始終保持一致。
- 筒倉/孤島保護主義：公司的數據情報隱藏於筒倉/孤島（silo），也是公司破壞者可以造成損害的地方。這些數據可能在財務部不知情下被隱藏，有關競爭對手的數據趨勢甚至未能與營銷或產品設計部門分享。原因很簡單：例如無知、懶惰、自我保護和自覺高人一等。
- 鏈接（Linking Out）魔術：使用逆向工程的第二個秘訣是鏈接。這是筒倉/孤島規範最難的部分，因為它與商業情資（不僅僅是數據智能）和商業戰略的深入理解程度有關。一位CEO曾公開介紹我是一名數據分析師/科學家，他的說法遭到我的糾正。我說：如果我只是一個數據分析師，你的項目會像許多其他項目一樣以失敗告終。我實際是一名商業情資顧問，也是一名數據分析師，負責創建業務結構完成任務。
鏈接的例子
那麼，鏈接是什麼意思？在逆向工程中，有一種侵入性和破壞性（invasive and destructive methodology）的方法來分析智能卡。用化學品一層又一層蝕刻智能卡：這種技術可以解開硬件和軟件的真實謎團。
同樣，通過分解滯留在筒倉/孤島區域內數據，能獲取數據智能增量收益，而這些正好是數據洞察力和隨後情資需要加強的地方。通過分解和鏈接，數據關係被鏈接起來，實際分析效果更加彰顯。如果不將數據抽絲剝繭，它將永遠隱藏起來，公司會像盲人騎瞎馬一樣不斷流失它的潛在價值。簡而言之，我們正在構建數據智能的規模和範圍。 迄今還沒有人明確指出這一創新技術會造成的後果。
最終遊戲：商業價值
最終遊戲總是意味着「商業價值」。就像《功夫熊貓》裏的阿寶爸爸：很簡單：就是用最好合適的的方法來煮麵條。
同樣，我們提取數據價值來為商業價值服務。如果不使用鏈接等方法，公司永遠無法在商業世界和市場競爭環境從相關數據的價值直接獲益。那麼我們所說的商業價值是什麼意思？
商業價值正在獲得新的增量收益並重新增加分析數據餅的大小。結合組織內的8個活動領域：它們是財務、成本、客戶、供應商、產品、效率、收入和基準。當我們認真審視業務戰略時，通常會涉及所有這8個活躍範疇。
然而，一旦我們進行數據智能化，就好像我們忘記了所有內在的核心領域。
來自數據智能的商業價值和數據價值是所有8項的總和。
終結遊戲：消除商業風險
要通過採用逆向工程實現商業價值的數據情資，我們必須將所有孤島視為對轉型成功同樣重要的部分。
基本上，它們都相互關聯的，並且它們是重新連接以形成相互依存結構的成分。這絕對是至關重要的。我們永遠不能在逆向工程過程中忽略某些部分或部門或過程：基本原理非常清楚。
簡而言之，如果其中一個構建塊或孤島存在數據缺陷並被破解，我們將對最終業務價值和結果產生連鎖反應。
基本原理是它們都是相互關聯的，它們是重新連接以形成相互依存結構的成分。這絕對是至關重要的。我們永遠不能在逆向工程過程中忽略某些部分、部門或過程：基本原理顯而易見。如果其中一個構建塊或孤島存在數據缺陷並被破解，我們將對最終業務價值和結果產生連鎖反應。
Data Intelligence and Reverse Engineering
The beauty of interweaving multi-disciplines is to open up new insights, broadening understanding of the issue in question. I learned so much when I was a postdoctoral fellow at John Kennedy School of Government Think Tank. Whenever we had discussions, we had multi-disciplines: atomic energy, political sciences, engineering, business restructuring, etc. Whatever the problem whether it came from corporations, governments, or charity organizations, etc. we, I can say, always came out with the best possible solutions. The current idea of STEM is in a sense, elementary as Sherlock Holmes will say: may be better described as the beginning.
To continue the story from the first article from the movie Kung Fu Panda, when the son asked his father: “What is the secret ingredient?” Indeed, one of the secret ingredients to data intelligence is Reverse Engineering.
1. Defining Reverse Engineering
Reverse Engineering is a process, to me it is also the mindset, the thinking and deductive reasoning approach to uncover with insights how a product, a system, a prototype, or a previous solution is developed and works, so we can uncover gaps, weaknesses in further improvement with linkages that were previously ignored, unavailable, or even not in existence. Now, just reread this sentence again and think. You will discover that this sentence gives us insights and actually the key to data intelligence. And this is the thrust and focus of this article: Data Intelligence and Reverse Engineering.
2. Leveraging Reverse Engineering
There is nothing new in leveraging Reverse Engineering, it has long been used with proof of concept for interfacing different software systems: such as Microsoft WindowsSCOPE, which shows the full contents of all files and APIs to enable the discovery of its competitive advantages and to restructure original documentation that has been lost or disappeared. Another great example is Bell Laboratories who used reverse reengineering to run on other machines than the original on the Apple Macintosh SE.
3. Reverse Engineering in Data Intelligence
So far, there has been no attempt in leveraging reverse engineering in data intelligence. With so many data projects, I have come to the conclusion by only asking one fundamental question: What is the issue?
Firms who pay for Data Solutions especially those for Digital Transformation Projects often fail. Why? Firms really want one thing, the Added Value. In practice this seldom happens.
Often, I uncover several different silos in existence (some not known to others) within the corporations. At the later part of the project, when crafting the solutions after uncovering so many data faults, I need to return to the original different silos.
The urgent desire to save all of us time is another reason why this article is being written. More importantly, I have worked out the process too.
4. Data Intelligence Made Simple
- The Problem – Most data consultants work from the Left towards the Right and never get to a Value End Point.
- The Solution – The secret is to work from Right back towards the Left so that the End and Start Points are always in harmony.
- The Silo Protectionism – A silo is where company data intelligence hides as well as where a company destroyer can create havoc. Data may be hidden without alerting the company finance department, or competitors’ data trends may not be shared with the marketing or product design departments. Reasons are often simple: ignorant, lazy, self-protection and status aggrandizement.
- Linking Out Magic – The second secret ingredient of using reverse engineering is to link out. This is the hardest part for silo specification because it involves a much deeper understanding of business intelligence (not purely data intelligence) and business strategy. Once a CEO introduced me as a data analyst/scientist, I corrected him even in public. If I am just a data analyst, your project will like many other projects, fail. I am a business intelligence as well as a data analyst who creates business structures for implementation success.
Linking Out Examples
- So, what do we mean by Linking Out? In reverse engineering, there is an invasive and destructive methodology in analyzing a smart card. Chemicals are used to etch layers and layers of a smart card: This technique can then reveal both the hardware and the software.
- Likewise, linking out is to make incremental gains by decomposing the stagnant silo zones where data insights and subsequently data intelligence gains should be. By decomposing and linking up, data is now linked up and the benefit of analysis actually increased. In short, we are building the size and scope of data intelligence. So far, no one has clearly pointed this technique and consequence out.
5. The End Game: Business Value
The End Game always meant business value. Like Po’s father in Kung Fu Panda: It is using the best method in cooking the noodles. In the same way, we extract data value to serve the business value. In short, without using such methodology as linking out, we can never achieve the company gains which are directly related to the way that it is operating in the real business world and the marketplace competitive environment. So, what do we mean by business value?
Business value is making new incremental gains and re-increasing the size of the analytical data pie. To combine the 8 active zones within the organization: they are Finance, Cost, Customers, Suppliers, Product, Efficiency, Revenue and Benchmarking. Often, when we examine business strategy diligently, we examine all these 8 active zones.
However, once we carry out data intelligence, it is as though, we forget all these inherent principal areas.
Business value from data intelligence and data value is a sum of all 8.
6. The End Game: Removing Business Risk
- To achieve data intelligence for business value by employing reverse engineering, we have to treat all the silos equally important to the transformation success.
- The rationale is that they are all inter-related and they are the ingredients reconnected to form a structure of interdependency. This is absolutely vital. We can never in the reverse engineering process ignore certain parts or departments or processes: The rationale is crystal clear. Simply put, if one of the building blocks or silos has data flaws and is cracked, we will have a knock-on effect on the end business value and result.
- Reengineering provides clear guidance to apply for uncovering the data intelligence that yields value, if not supremacy for your business in a competitive environment.
