Genpact recently hosted an advisor/analyst session at its new Innovation Center in Palo Alto. So why is a BPO specialist opening a center in Silicon Valley? Genpact is using the center as a hub for the development (and showcasing) of platform-based digital assets aimed at mimicking the industry disruptors by being based on standardized, simplified operating models and distributed technology that can be deployed at web scale as utilities. The center is also being used to house design thinking (DT) workshops, and as a co-innovation Lab.

Genpact views its role as bringing domain knowledge and understanding of process and then leveraging DT and digital technologies, and emphasizes it is prepared to destroy existing BPO revenues in the process. It recognizes that operating to traditional upper quartile and best-in-class standards is no longer adequate as organizations look to compete with new forms of digitally-based competition. The emphasis in Genpact’s digital strategy is to continue to focus on middle- and back-office processes but to reorganize these processes with greater emphasis on straight through processing (STP) and on real time insights. So, while Genpact continues to put process before technology, the company is focusing on the application of 12 key technologies, specifically cloud/SaaS, mobility, dynamic workflow, advanced visualization, RPA, machine learning, cognitive computing, NLP, NLG, IoT, data analytics, and autonomic computing.

In general, Genpact is partnering with, rather than acquiring, companies with these technologies, though it has acquired PNMsoft due to the critical importance of dynamic workflow technology as a backbone for new digital processes, and it also continues to invest in IT services companies with expertise in applying these technologies, hence its acquisition of Endeavour Technologies to strengthen its mobile capability. Key technology partners who attended the session are:

• PNMsoft, an Israel-headquartered provider of workflow, BPM and case management (HotOperations) software and Microsoft Gold partner, recently acquired by Genpact.  Though Genpact does not yet have experience of using PNMsoft solutions with a client, it will have been attracted by the ability of software such as PNMsoft Sequence to enable organization to establish multiple versions of workflows for a particular process, e.g. loan origination, and judge the impact of moving work between each of these workflows in terms of process SLAs and cost. This approach also means organizes can test multiple workflows and change/optimize processes without service interruption
• Automation Anywhere, combining cognitive, analytics, and RPA technology
• Rage Frameworks, developing “intelligent machines” for a platform-based approach to knowledge work. Seven of the current 16 “intelligent machine” platforms being developed by Rage Frameworks are being developed in conjunction with Genpact
• Arria, a small U.K. based Natural Language Generation specialist.

Genpact is increasingly emphasizing its role in assisting organizations in creating new digital business models, and is looking to build digital assets based on combinations of the 12 technologies. Pilots and live examples demonstrated at the Innovation Center included:

• Insurance policy servicing automation. Here the utility service on offer takes policy change feeds from a variety of channels, potentially including portals, emails, contact centers via use of voice-to-text technology, and uses NLP technology to identify appropriate data and update the policy administration system (PAS). Where exceptions occur, it can send details to an agent or request more information automatically, generating a NLG response to the originator. This utility is currently live at two Genpact insurance BPS clients
• Using neural chat in support of opening corporate banking accounts. This incorporates use of OCR technology to extract data from images of documents taken via smartphone, using neural chat to present details of missing information to both the agent and the customer simultaneously for validation
• Wind turbine predictive maintenance. In this example, live since Q3 2015, sensor data is being used to monitor a fleet of 5,000 wind turbines to identify the failure rate of parts including the likelihood of part failure in a particular location and month (taking into account the age of parts, the location of the part, and the impact of the weather). This data is then combined with information from the ERP system to identify the potential cost in terms of lost production and wider damage of a part failure, together with the cost of holding individual parts to optimize the number of parts held by individual location. It has enabled the client to move away from scheduled seasonal maintenance to a more predictive response to maintenance. The client has benefited from increased uptime (more revenues) and a significant reduction in maintenance costs. This approach is clearly applicable in areas such as cable network management and aircraft maintenance
• Use of NLG in support of management reporting to produce a commentary on the underlying data and charts, aiming to produce one version of the truth and focusing the interpretation of the data and graphs onto the most salient points, thereby avoiding individual managers coming to differing conclusions from the same data
• Reimagination of the quotes process for an electrical distributor. In the current process, the distributor receives electrical blueprints from which it manually extracts SKU-based orders to enter into its order management system. This is typically taking 3-hours per blueprint. Genpact has developed a pilot incorporating computer vision to input the blueprint, NLP to identify table data within the image, and machine learning to enhance the identification and interpretation of table data, which can then be input into the order management system. In this instance, the supervised learning exercise in support of the machine learning involved ~12 personnel over a 6-week period. This example, which is not yet live, illustrates Genpact looking at areas of a client’s operations that could be completely digitized; its proposed solution eliminates all manual processing.

Some of these examples are not yet in production, but all are evidently transformational. Clearly, Genpact is at an early stage in its development of digital assets, and the technologies and the technology vendors with which it works will evolve considerably. But it is clearly investing in Lean Digital Innovation in earnest: with examples such as the reimagination of the quotes process for an electrical distributor, Genpact has come up with an offering that is very different from its legacy in BPO services: presumably the commercial model, which I was told has yet to be finalized, will be transaction-based.

Among the strengths of Genpact’s approach with Lean Digital are its:

• Emphasis on using combinations of emerging technologies rather than on single technologies to create point-based digital assets
• Domain knowledge and understanding of process (its heritage): the approach starts with a focus on transforming a business process rather than with the application of automation/AI
• Evident understanding that emerging technology companies just want to try some pilots with real clients rather than spend time on contractual arrangements and joint marketing.

********************

See also Genpact Assists Client in Targeting 10x Process Improvement, Applying Design Thinking to Order Management by John Willmott, published this week here.