ENGIE is a French utility company specializing in energy transition, electricity generation and distribution, natural gas, nuclear energy, renewable energy, and petroleum. As a global player in low-carbon energy and services, ENGIE is one of the world's largest independent power producers, with annual revenues of US$ 60 billion. Its network is comprised of 170,000 employees, customers, partners and stakeholders, with operations in 70 countries across five continents, and over 8.5 million retail customers under its consumer utility operations unit.
In today’s digitized world, utility service providers such as ENGIE rely on hyper-automation to ensure the smooth operation of functions such as the compilation of meter readings, invoice generation, payment processing, and customer request processing. These automated tasks, executed in a timely manner through a series of associated IT tasks, or batch processes, ensure business continuity. However, any disruptions, failures or delays in processing can adversely affect operations, leading to potential revenue loss and less than optimal customer experiences.
With millions of customers accounting for over 12 million utility contracts, varying billing cycles and meter connection types, ENGIE conducts two million meter readings and generates over 150,000 invoices each day. To execute on such large-scale requirements accurately and on time is a huge operational undertaking, involving a complex series of workloads to draw massive amounts of data agnostically from various applications to update records, generate invoices, and tag incorrect invoices and bills. To manage such a complex IT landscape, ENGIE employs a Linux-based scheduler to orchestrate the automation of workload across its complex IT infrastructure, involving a wide range of applications such as Customer Relationship Management (CRM), billing engines, financial applications, printer vendor systems, and various ETL[1] tools and file systems. Since the batch processes serve critical business needs, ENGIE needs to monitor and detect issues, diagnose them, and resolve them quickly.
Identifying the challenges
ENGIE was experiencing operational challenges in terms of incorrect invoicing, missed payment processing and inconsistent payment records. For a global enterprise with millions of customers generating transactions each day on a huge scale, even a minor system issue with billing or invoicing could lead to significant negative business impacts. ENGIE identified two main areas within its incumbent IT infrastructure that were proving to be problematic:
(1) Lack of autonomous monitoring and intelligent resolution
ENGIE’s IT system was not capable of automated monitoring or forecasting, instead relying heavily on manual monitoring. The company’s IT team was responsible for monitoring every application, within the Linux scheduler as well as in the SAP ERP system, to ensure that all processes were completed. This obviously required extensive efforts and resources, which were additionally hampered as processes could not be monitored manually during non-business hours.
Manual monitoring often resulted in lags in identifying anomalies, or in diagnosing and resolving issues, resulting in serious delays in business operations. For example, delays in accurate digital capture of meter readings hampered the process of records against the customer’s account. This caused slippage in calculating charges and sending out bills, which led to delayed revenue realization. Additionally, it often resulted in inaccurate charges, the generation of duplicate copies of bills, and incorrect communication with customers, all of which led to an increase in customer complaints, in addition to millions of Euros in unbilled and incorrectly billed invoices. To compound these issues, the existing Linux-based scheduler was unable to store historical data in a usable format and or integrate with external systems to provide real-time data. This proved to be a major obstacle in generating analytics that could have been used to identify, predict, and avoid such IT failures.
(2) Reactive workload management and no defined SLAs
ENGIE’s workload management was geared reactively rather than proactively, relying on pre-defined rules to generate alerts in cases of process failure, rather than live contextual data. In addition, there were no service level agreements (SLAs) for any processes, so operational parameters were not in place to define optimal timelines for process start time, execution time, and completion time. The IT platform had no mechanism to generate alerts or notifications, which meant that if any component malfunctioned, IT personnel would only realize it after those key processes stopped.
The absence of planning often meant that process completions were allowed to drift, leading to escalation requiring costly manual intervention, as well as business process violations and/or poor customer experience. For example, when receiving payment information from third-party payment gateways, customer payments are routed through middleware to ENGIE’s SAP ERP system for processing, and then the amount is credited to corresponding accounts. This entire process is done by batch jobs that execute at the close of each business day. If that process doesn’t complete in a timely manner, it can result in the loss of data or inconsistency in data, leading to faulty payment reminder processes. This causes unnecessary inconvenience for the customers and increases dissatisfaction and complaints.
ENGIE’s digital transformation journey
Having identified the technology challenges it faced, ENGIE needed to reimagine its IT infrastructure through digital transformation. This required the integration of a solution that could not only monitor workload automation processes across its extensive IT infrastructure and business applications to reduce dependency on manual issue resolution, but could define SLAs, proactively identify potential SLA misses, and curb any impact on business operations to prevent historic issues from occurring. The automation solution would be required to help ENGIE achieve several key business goals:
- Reduce revenue loss or delay by addressing backlog issues, such as unsettled accounts, that have a direct impact on revenue realization.
- Prevent incorrect billing and invoice generation while reducing payment transaction errors, as incorrect invoicing and inconsistent payments can cause major inconveniences for customers and risk statutory penalties.
- Improve customer satisfaction by making the shift from reactive to proactive CRM and problem-solving, working to meet customer demands.
- Enable cross-application synchronization to ensure 100% accurate billing and invoice generation. This would require integrating CRM and billing services with entities such as distributors, banks, financial agencies and numerous vendors and service providers.
- Better revenue assurance from autonomous monitoring and intelligent resolutions.
- Improve SLA compliance by leveraging predictive analytics.
Integrating ignio
ENGIE partnered with Digitate, a SaaS-based autonomous enterprise software provider, to embark on its digital transformation journey to address its legacy business challenges with a focus on improving operational efficiency and accuracy enterprise wide. Incorporating Digitate’s flagship ignio AIOps into its operations, ENGIE implemented a structured approach to automation in order to turn business activities into intelligent processes.
ignio is distinct from other AI-driven platforms as it offers closed-loop last-mile automation. To achieve this, ignio leverages a three-tier mitigation approach – continuous monitoring, proactive prevention, and reactive correction – to reduce business impacts. Once the pattern for a problem is identified, proactive and reactive mechanisms work to eliminate approximately 90% of the impacted volume.
In ENGIE’s case, when a potential issue or risk was identified, a corresponding use case was brought in as a preventive mechanism to remediate issues effectively. The correction in bulk was executed through ignio automation to rectify and optimize operations at scale. ignio was also able to extend to ENGIE’s CRM system through a custom CRM solution over the database layer in the infrastructure, enabling ignio to integrate and synchronize agnostically with any system, regardless of the specific applications involved.
Digitate identified four key areas to deploy automation: service request operations, customer experience, cross-application synchronization, and cognitive alerts. Using ignio provided ENGIE with the ability to enhance its automation capabilities across service requests, incident management, batch job management, unbilled revenue resolutions and systems health checks. Digitate deployed 56 out-of-the-box use cases with ENGIE to leverage AI and machine learning to streamline business processes. In addition, ignio AI.ERPOps boosted automation capabilities across several key functions for ENGIE, including the validation of meter reads, bill calculation, streamlining of payment postings and refunds, and blocking of erroneous late penalties and incorrect invoice amounts.
Relating back to the two main problem areas that ENGIE had identified, the following top-level IT goals were achieved:
(1) Better revenue assurance from autonomous monitoring and intelligent resolutions
Complete visibility into the IT landscape is the key to better monitoring and with that in mind, Digitate provided ENGIE with a multi-layered solution for monitoring workload processes, enabling a digital blueprint to be created of ENGIE’s entire batch system, reflecting the jobs, their schedules, interdependencies with each other, and their historical behavior. Data was agnostically collected from a variety of sources including execution logs, databases, and email reports, to ensure the platform has access to both real-time and historical information. This blueprint helped the ENGIE team understand the normal behavior of any job by accurately capturing process flows in depth, identifying focal areas of concern, pinpointing issues in progress, and leveraging the benefit of historic context tracked with current data. The net results? The ENGIE team can spend less time and energy on manual monitoring because they have always-on autonomous monitoring of the entire workload ecosystem and timely notifications of any potential breach.
ENGIE saw a reduction in batch-induced delays and risk of batch failure, lower cost of monitoring, reduced effort and time to eliminate batch issues and an improvement in its competitive position due to higher customer satisfaction. In addition, the closed-loop ignio model enabled ENGIE to diagnose the probable root cause of anomalies, analyze the impact on critical business operations, and recommend the right solution to rectify it, helping ensure business continuity.
In terms of tangible results, ENGIE saw an 80% reduction in impacts to downstream processes like billing or payment, a more than 95% reduction in customer complaint tickets, and an average of around $5 million per day of revenue loss prevented.
(2) Improved SLA compliance by leveraging predictive analytics
ENGIE also leveraged to uplevel its workload management processes from reactive to proactive. Drawing on knowledge of the existing workload ecosystem and historical run data, ENGIE uses ignio to identify performance trends and patterns and derive optimum thresholds for every job run. Leveraging AI provides ENGIE with dynamic recommendations for SLAs for the execution and completion of business-critical operations. These SLAs can adjust themselves relative to the operating environment changes to ensure day-to-day requirements are met, prioritizing focus on alerts that are important and meaningful, and reducing noise.
One of the biggest factors enabled by deploying ignio is the ability to make real-time predictions that could detect potential SLA breaches several hours ahead of time, diagnosing and localizing their probable causes, and recommending fixes. This has enabled the prevention of many outages and minimized the impact of others, enabling ENGIE to become more proactive in its IT operations (ITOps) process. For example, for customer payments, ignio analyzes the data captured from the payment gateways and payment files in real-time to predict if the payment will be successfully reflected or not, as per the defined SLAs. In case of any discrepancies from normal behavior, it assesses the impact on the subsequent business operations and accordingly informs the downstream applications to send or hold dunning letters and reminders to the customer accounts. This increases customer satisfaction, as they do not receive unnecessary or inaccurate reminders, and ENGIE’s profitability, as payments are reflected in a timely manner.
The operational benefits for ENGIE include better enforced SLAs, the implementation of recommended thresholds for business SLAs, improved customer experience, reduced alert noise, and lower risk of missing alerts. ENGIE is now able to leverage AI to predict around 90% of SLA misses and is able to look two to three hours ahead of time for any business SLA failure, substantially improving the ability to predict and prevent failures through corrective proactivity.
Overall impact
The results of the Digitate implementation for ENGIE include near zero incorrect billing and invoice issues, improved resiliency and profitability through prevention of revenue leakage, and substantial reduction in front desk inquiries, improving customer experience and removing frustrations for both customers and the ENGIE team. The 18-month project paid for itself within the first month of its launch, and to date has delivered the following specific outcomes:
- €4.2 million (US$ 4.4 million)in revenue realization.
- 84% service request automation.
- 95% MTTR (mean time to resolve) reduction.
- 100% reduction in downstream business impact.
- 90%+ unbilled account service request tickets handled by ignio.
- 150,000 bills per day generated with near zero issues.
- 8000 new customer invoice issues were mitigated in just one week.
- 40,000 incorrect customer correspondence items blocked.
Ultimately, ENGIE’s digital transformation journey to automate service requests and business processes has enabled them to resolve issues faster and realize improved operations, lowered costs and higher customer satisfaction and retention.
[1] Extract, Transform and Load, a three-step data management process that extracts data from both structured and unstructured data sources, transforms it into an acceptable format for requirements of the business, and sends to a target destination.
As Vice President, Sales and Client Services, Rajiv Nayan is responsible for customer delight and revenue growth for Digitate. Rajiv has an accomplished career of more than 20 years in enterprise software and IT services. He graduated from Delhi Technical University and earned an MBA, with a specialization in finance and strategy, from the Rotman School of Management in Toronto. He then joined IBM Canada as part of its executive leadership program. Before joining TCS, Rajiv was responsible for business growth and delivery as Industry Cluster Leader for TCS Canada. He has also worked at GE and JP Morgan Chase. Rajiv is passionate about investment in start-ups, running marathons, and promoting social causes. He is a board member of several nonprofits working on sustainability and cross-cultural understanding.
