On March 1, we announced new capabilities powered by generative AI. Our Chief Product Officer at Pega, Kerim Akgonul, went on to expand on this announcement, detailing why we think generative AI technologies like ChatGPT will revolutionize all facets of business and how this helps accelerate the path to the autonomous enterprise.
All the announced capabilities were showcased at PegaWorld iNspire 2023, where we dove deep into generative AI. We explored:
- How
AI will transform business in the autonomous enterprise keynote - How to get value from AI in Aflac’s keynote
- How to prepare – with an eye to responsibility and governance in the "Mapping the AI Landscape from Generative, Predictive to Prescriptive AI" and "Redefining the A in AI" breakout sessions
- What AI looks like in action – with live demos of generative AI on the Innovation Hub floor
What could generative AI mean for back-office operations?
To date, leaders have started to explore and adopt these technologies largely for the purposes of customer engagement: generating marketing copy for ads, emails, and chats (which is also a capability Pega announced on March 1).
But fewer businesses have started to explore the use of these technologies in the back office, even though the back office is ripe with areas to improve. Operational work is complex by nature:
- End-to-end customer journeys traverse multiple workflows, teams, and apps
- Data lives across multiple sources
- Rules and regulations differ based on customer needs, region, status, etc
This leaves back-office employees in a spot where they need to read and interpret a lot of data, reports, documents, and text to get work done – slowing them down, leading to potential errors, and causing a lot of frustration for them and their customers.
So while there will be millions of opportunities to use generative AI in the back office, here are five potential applications focused on helping reduce manual work and enabling better decisions with fast, concise operational insight.
#1 On-the-fly insight into operational KPIs
Visibility into what’s going on across end-to-end workflows is critical for ensuring operations are smooth, efficient, and issue-free. Back-office leaders want fast insight into operational metrics – answering questions like:
- “By type, how much work is assigned to each of my teams?”
- “By region, how quickly is work getting done?”
- “What stage of the process is slowing us down the most?”
At a lot of organizations, creating and maintaining dashboards and reports to answer these questions is time consuming – often requiring a lot of back and forth between data analysts and the operational leader looking for insight.
Pega Explore Data is a fast way to get insight from workflow data – enabling users of workflow apps to directly slice-and-dice operational data on-the-fly for shareable charts, reports, and dashboards.
Generative AI will make it so business leaders’ questions can be answered immediately – by turning those questions into automatically generated charts, reports, and dashboards.
A single employee in a large-scale back-office operation might have 12+ applications in which they have assignments in, totaling 100+ tasks or more. Without quick insight, they could be left lost when trying to figure out what’s on their plate and where they should spend their time.
Pega Process Fabric® gives employees a single launchpad into all of their workflows. It listens to all the work getting created and updated across all workflow systems (Pega & non-Pega) to give:
- Each individual a single, prioritized worklist of everything they need to get done.
- Managers and back-office leaders visibility into how work is progressing across their teams and workflows.
Generative AI can help turn operational insight in a tool like Process Fabric, into a concise, conversational summary of what’s going on. Reports – along with headlines, descriptions, and data from all cases – could be pumped into a generative AI model to produce a four bullet conversational summary for each individual with: what’s on my plate, what’s pressing, what’s new/updated, and what should I work on first.
When an employee opens a new assignment to get work done, there can be a lot of context to absorb in order to get up to speed. Before making any decisions, you need to quickly understand things like:
- Where did it come from?
- What steps have been taken already?
- What notes have been added by my teammates?
And more.
Case history timelines help give a quick visual assessment of what steps have been taken for work, but that still leaves an employee to read through notes and descriptions to get full context.
Generative AI is great at summarizing lengthy text and even at distilling logs to provide a short, consumable summary – and has the potential to augment timelines and history into a short conversational narrative.
If you’ve ever emailed a large organization to get something done, you know that responses can be slow (if they ever come at all). That’s because often, someone on the other end is reading, interpreting, and triaging all emails. Sending them to the right place in the back office to get something done. Then, once the work is complete, that same person is writing and sending a response your way.
AI-powered capabilities like Pega Email Bot help drastically accelerate the turnaround for customer emails – even at high volume.
Email bot understands the topic of incoming customer emails with natural language processing (NLP) and automates the back-office workflows. It then responds to customers automatically with the appropriate template… assuming that template exists.
Where that leaves organizations, is that they can either:
- Spend lots of time and money creating and maintaining templates for their customer emails and chats
- Or, send boilerplate customer emails – which sound impersonal and can lack the details the customer is looking for
Similar to how businesses are beginning to adopt generative AI to automate marketing content generation, it can play a huge role in generating customer correspondences. Initially – it can be used to quickly generate and regenerate templates – reducing the burden on employees to sit down and write them. And in the future, as model controls improve, there will be an opportunity for generative AI to produce or augment direct customer responses on-the-fly – greatly reducing reliance on templated correspondence.
According to a McKinsey report, employees spend over nine hours a week searching for information to do their job – which includes reading through lengthy documents. And the number can be much higher in document-heavy functions like compliance, HR, tax, claims, and legal.
Intelligent document processing can help extract structured information from semi-structured documents, but requires model training, fine-tuning, and maintenance to perfect.
AI has been used to summarize news articles for years. The same approach can be applied to lengthy in-house documents like insurance, tax, legal, or HR forms. Generative AI can ingest text and provide condensed versions, deliver key bullet points, and extract structured data points like locations and names.
The possibilities are endless
These ideas are just the tip of the iceberg. Generative AI will transform the future of business across all areas, and back office is no exception.
Pega is committed to both simplifying back-office operations and innovating with generative AI. Learn more about Pega's GenAI Capabilities today.