Robotic Process Automation
Robotic Process Automation is gaining traction
Robotic Process Automation (RPA) has attracted considerable attention as a way to automate repetitive clerical tasks, by mimicking the way human workers carry them out. Since the introduction of the term (around the year 2000), RPA has evolved from simple screen scraping and desktop automation to the promise of Cognitive RPA. Reports by industry analysis leaders estimate the global spending on RPA software to reach $2.4B in 2022, with annual growth rates over 50%.
While the RoI of these investments is quite apparent, most analysts also stress that automation does not necessarily imply intelligence. In a recent article published by Forbes (“Sorry, but your bots are stupid”), Ron Schmelzer stresses the fact that automation is inherently dumb, and that automated software bots are still dumb. Concluding that “despite much of the marketing hype, what is being sold as intelligent automation is far from intelligent.”
Towards Intelligent RPA
If RPA (or BPM, Business Process Management, its more strategically-oriented elder sibling) is going to succeed in boosting productivity in the coming years, this will be mainly dependent upon the integration of cognitive components showing higher degrees of intelligence, and, in most cases, keeping humans in the loop.
For instance, extracting information from unstructured information sources is a common task in myriads of business scenarios, such as:
- On-boarding processes (data and documents required from new clients or employees)
- Claims and complaints management
- Risk analysis from financial reports
- Voice of the customer analytics from any interaction channel
- Helpdesk and contact center optimization and quality audit
- Compliance monitoring and audit
- Due diligence processes
- Cybersecurity
- Social listening
- Pharmacovigilance
- Selection of candidates for clinical trials
In the past, all the aforementioned processes have been labor-intensive and obvious candidates for BPO (Business Process Outsourcing). Nowadays, however, many of them are still waiting for intelligent automation.
Natural Language Processing in RPA
Natural Language Processing (also known as Text Analytics) solutions are mature enough to solve typical insight extraction problems in the scenarios outlined above, conveniently and cost-effectively. The classification of complex documents (or their sections) and the classification of interactions (e.g., via contact center or helpdesk), as well as, insight extraction with semantic approaches is the natural choice under the following circumstances:
- When no training set is available.
- When high accuracy is a must.
- For small data / sparse data.
- When your analysts’ time is precious (and expensive)
- From a few clues provided by experts.
And last, but not least, Natural Language Processing (NLP) and Machine Learning (ML) techniques are not competing but complementary in many business situations. MeaningCloud adopts the best of both Artificial Intelligence (AI) approaches to address intelligent automation for the benefit of our clients, integrated with RPA/BPM software suites. Learn more in this post.
Take a look at our Text Analytics APIs and integrations. Give us a try!
