It’s a well-known estimate that between 80% to 90% of organizational data is unstructured, or what we call dark data. It’s all that information from both physical or electronic documents, emails, PDFs, chats and files, some of which is valuable for illuminating trends and dictating strategy — if only we could find the key to unlock it. Businesses have been tackling this question for years. Face it, there are people you can hire to input much of this data, but that’s time-consuming, expensive and results are delayed.
On the other hand, there’s all this hype around artificial intelligence and machine learning. The hype is real and can boost the value of your data, with the right code, architecture and understanding of how humans complete tasks and processes. Understanding the decisions people make, what data is available and the multiple sources and discrete systems that can be mined is one step towards gathering the context digital workers will need to succeed.
But getting the context is just half the equation. What about the rules that govern a process? All too often, the business rules are locked away in process documents and application rules (i.e. application code) or based on tribal knowledge. By understanding how these factors work together and the processes behind them, organizations can automate them using AI and machine learning. An autonomous future has been born — but only if we can adapt to the new technology using and sourcing context to amplify meaning, and therefore, amplifying value.
In the autonomous future, business tasks will be completed by AI bots and algorithms. Complex processes will be accomplished by an ensemble of bots and systems working in harmony. The future system will be driven by context and will need access to the rules that governed the processes previously.
There are two approaches to progressing the autonomous enterprise.
- We can build digital workers that have business logic baked in and have been trained on the rules they are to follow. This is where RPA is today.
- We can build truly autonomous digital workers that are trained as part of a larger process but get their rules or directions at runtime
The problem with the first scenario is these digital workers struggle to adapt as the business evolves. The bots are embedded with business logic, which requires constant maintenance and tuning to stay abreast of the evolving business ecosystem. Embedding business logic in user interfaces has been an anti-pattern in software development for many years but it’s exactly what we’re doing with the current RPA strategy.
The future of business autonomy lies in the second strategy and will require that we begin to create digital workers that consume both the context and business directions at run time. For the second scenario to become a reality, organizations will need to be able to codify both context and business directions into a new type of digital contract. These digital contracts will fuel the context-aware digital workers. But what exactly is this digital contract?
The term “smart contract” was first used by Nick Szabo, a computer scientist, scholar and cryptographer, to describe the concept he envisioned using computer code to codify the terms of a contract. The smart contract contains the rules under which the parties of the smart contract agree to interact with each other. When the predefined rules are met, the agreement can be enforced. The challenge with this original idea is that many times the parties of the agreement do not have access to the data or context they need to verify if the rules or conditions have been met. For smart contracts to be an enabler of enterprise automation, we must expand what gets included in smart contracts to include the business context and the rules.
However, codifying business context and directions into digital contracts isn’t enough. We must get the data to the digital workers and we must do it in a secure, trusted and auditable way. To accomplish this, digital contracts need to be published into a secure and trusted business partner ecosystem where digital workers operate autonomously. This trusted partner ecosystem must provide data sharing via advanced encryption of data, integrity, authenticity and privacy, low transaction costs and near real-time information exchange. While smart contracts were originally envisioned to work as part of a distributed ledger or blockchain technology, the jury is still out on whether the typical enterprise will need the added complexity that comes with a distributed ledger. The bottom line is that we need a trusted environment where machines are free to commit business transactions under the terms of immutable digital contracts.
Moving beyond data capture and enriching your data into a business-critical context to fuel digital workers will lead the way to extract value from your data. Building better smart contracts and creating trusted technology will advance the race to more autonomous processing. Only then will efficiency, productivity and profitability follow.