Future of Financial Services Workforce

UntitledFinTech disruptors have been finding a way in by focusing on a particular innovative technology or process in everything from mobile payments to insurance. A forte of technologies “AI-ML-DL-NLP-CV” is fueling the FinTech innovations. The large financial services companies can’t be complacent as FinTechs have been attacking some of the most profitable elements of the value chain and as well as areas which were historically subsidized.

Let us refresh our memory on these AI technologies and their relevance to the financial services industry.

  • AI makes machines to learn from experience and perform human-like tasks – AI offers robotic & intelligent process automation (RPA/IPA) of financial processes
  • ML is a specific subset of AI that trains a machine on how to learn – ML is enabling algorithmic trading lead to better predictability and decisions around credit and consumer lending, thereby lowering risk to the bank or financial institution
  • DL is s a type of ML that trains a computer to perform human-like tasks, such as identifying images – leverage big data (customer demographics, consumption records, etc.) to parameterize a DL model that can simulate the likely response to new product/service configurations (e.g. new credit card with cash rewards, moderate interest, zero interest on balance transfers, etc.)
  • NLP is a branch of AI that helps computers understand, interpret and manipulate human language – NLP is shaping the future of banking with voice assistants and ubiquitous computing.
  • CV s a field of AI that trains computers to interpret and better understand the visual world –  CV is transforming financial services by using appealing visuals and new solutions for a new world where seeing is believing

These new-age FinTech developments are leading to a continuous transformation of the financial services workforce. The changing landscape and evolving financial services resource pyramid is presented in the diagram above. I would like to highlight a few trends reshaping the talent of financial services on this blog post.

  • AI automating business-as-usual activities of financial services: Robots and AI already started addressing key pressure points, reduce costs and mitigate risks. Building capabilities to target a specific combination of capabilities such as social and emotional intelligence, natural language processing, logical reasoning, identification of patterns and self-supervised learning, physical sensors, mobility, navigation and more are in swing. The goal is to look far beyond replacing the bank teller. There are whole categories of work that had not been seen as cost effective to automate. However, with lightweight software ‘bots’, workers are freed up to focus on higher value activities.
  • Changing patterns with Human vs Machines foray: Are financial services firms moving to re-shoring of work with talented machines? The answer seems to be, Yes. In the last two decades, many financial firms have ‘offshored’ repetitive tasks to lower-cost locations such as India, China, and Poland. However, relative costs for labor in those regions have started to rise. Combine this with improvements in robotics and AI capabilities and machines are becoming credible substitutes for many human workers. As the capabilities continue to improve and technology continues to drive down the cost of machines, these forces will combine to spur re-shoring, as more tasks can now be performed at a competitive cost on-shore. Even functions that seem dependent on human input, such as product design, fraud prevention, and underwriting, will be affected. At the same time, the need for software engineering talent will continue to expand
  • It is not just automation, Technology is picking high-end work: ML is enabling next-generation algorithmic trading systems are moving from descriptive and predictive to prescriptive analysis, improving their ability to anticipate and respond to emerging trends. And while algorithm trading programs were once limited to hedge funds and institutional investors, private investors can now get access to them too. AI soon automate a considerable amount of underwriting, especially in mature markets where data is readily available. Even in situations where AI does not completely replace an underwriter, greater automation would allow humans to concentrate on assessing and pricing risks in the less data-rich emerging markets. It would also free up underwriters to provide more risk management, product development advice and other higher value support for clients.
  • While building machines, the real focus is on accessing the necessary talent and skills to execute strategies and win markets: Financial services firms lack the internal knowledge and expertise need to implement a customer-centric approach. For example, a mainframe programmer who maintains a core banking platform may not have the skills or interests to learn to code AI applications. Many senior IT executives, non-IT staff-members, and even technical personnel do not have the skills needed to build and operate an effective digital channel offering. Financial institutions are starting to realize they will need talent with very different skills. This might mean finding more industrial engineers for robotics work, or retraining underwriters to do higher value work once AI is used to automate certain existing functions. But the issue runs deeper than developing a different competency model. First, firms to understand what is already working and what needs to be done differently. This might involve changes across the human capital strategy through revitalized recruitment, learning and development, partnering and cultural initiatives.
  • The contingent workforce is creating the talent-exchange mindset: financial firms need to address is the growing preference for flexibility and entrepreneurship among many in the labor force. In the United States, the US Chamber of Commerce has found that 27% of the labor force is currently self-employed, and some believe that this ‘contingent workforce’ could rise to 40% or more within several years. Practically, for this reason alone, financial institutions will need to adopt a ‘talent exchange’ mindset, leveraging part-time and/or self-employed individuals in a creative manner. This may range from bidding out specific tasks or work to expanding the use of seasonal or temporary workers. Of course, this will introduce challenges around culture and quality, and this will introduce new opportunities as well. For example, we might see employers using online platforms to manage confidentiality and legal risks in creative ways.

Artificial Intelligence capabilities impacting the financial industry and thereby attitudes toward work continue to change, some of the attributes that have benefitted institutions in the past such as big firm and stable employment are slowly losing their appeal. Refreshing financial firm’s approach to recruiting, learning and development, and culture may offer an effective way to address issues that FinTech has brought into the open market.

Welcome your ideas in further spotting future trends in financial services workforce.

 

Internet of Medical Things (IMoT)

imot

Many healthcare firms and consumers have latched onto the Internet of Medical Things (IMoT) by way of wearables, such as FitBits and Garmin watches, referred to as “FitTech.” With over 2/3 of medical devices estimated to be connected over the next 3 years, IMoT is going to have a significant impact in Healthcare operational and financial processes. Let us examine the impact of IMoT on healthcare payers, health providers, and consumers.

A) IMoT and Payers: 

i) Underwriting: The first process comes to mind is Underwriting. By equipping consumers with IoT-enabled medical devices underwriters can better understand what an individual’s health looks like daily, rather than at long historical intervals. With this wealth of information at the helm for an individual, underwriters gain access to health data from periods of time that used to be non-existent in health records and claims. IMoT can enable underwriting for

  • improved bottom-line of the payer by better understanding what each individual new member will cost them
  • increased wallet share by preventing lower-risk members from being improperly marked as high-risk based on one-off health encounters
  • Optimized underwriters’ time spent on due diligence, especially avoiding unnecessary full medical underwriting (FMU.)

ii) Preventive Care: Preventative care is a perfect application of IoMT. Biometric sensors and other devices can collect real-time data from health plan members, help point to higher-risk metrics or lifestyle choices, and notify payers to get the correct members enrolled in prevention programs. Oe successful use case is Beam Dental. The Beam Brush tracks an individual’s tooth brushing habits (such as the dental habits of employees under their employer’s insurance plan) and allows their good habits to drive down the cost of dental insurance for their group. By activating members to take control of their health before chronic or acute health issues arise, payers will see success in loss prevention as well as a happier (and healthier!) member base.

iii) Claims and billing efficiencies: IoT can aid in cumbersome tasks that waste administrative hours by leveraging AI led solutions, such as determining whether a claim should be accepted or rejected for minor claims or processing payments. By freeing up administrative time from these tasks that can be automated, payers can invest more in programming for their members to drive focus towards prevention leading to savings in administrative costs and savings in claims payments from healthier members.

B) IMoT and Providers:

IMoT has the potential to facilitate remote patient care to optimizing hospital operations to streamlining data management, healthcare providers can leverage the lucrative potential of IoT. These use cases are elaborated below.

i) Improved hospital operations: IMoT can be introduced and ramped up to optimize a hospital’s daily functions and cut unnecessary costs. Tracking medical assets within a facility is a good use case. Every year, millions of dollars bleed from hospitals from lost or stolen equipment. By attaching sensors (e.g., RFID or Bluetooth) to equipment, hospital staff can track the exact locations at any point in time, allowing for better oversight. This can solve the problem of lost equipment, reduce theft, and even track overall use of equipment. The life of medical equipment varies greatly based on the frequency of use. By tracking movement over the life of a piece of equipment, hospital administration can get a better idea of when to replace or schedule maintenance to avoid periods of time where equipment is unusable.

A second IMoT use case is in intake or discharge processes. With IoT, unobtrusive sensors can be placed in patient wristbands and staff badges better to track how quickly patients flow through different areas of the hospital (such as pre-op rooms to the operating room) or how efficiently staff attends to a given patient. This can remove backup from current bottlenecks in flow at the hospital, including but not limited to Emergency Department wait times, intake, discharge, and shift changes.

ii) Interoperability and Data Monetization: IMoT at a basic level improves existing systems for providers. For example, biometric devices and sensors are often system-agnostic and can connect through APIs to multitudes of EHR systems. If a patient has doctors in multiple health systems, their disparate EHRs (and therefore doctors and care plans) can be updated accordingly. The IMoT combined with AI/ML and NLP can nurture the massive loads of HC data. By relying on IoT-enabled technologies, providers will no longer deal with unusable, unstructured data but rather well-organized and insightful data systems. The world of well-managed data in hospitals and health systems opens up with the adoption of forward-thinking technology. Doctors can better tailor care plans to patients’ specific needs based on historical data of like patients and avoid oversight of potential complications such as contraindications.

iii) Expanding remote care revenue streams: IoMT eases the implementation of remote patient care. With IMoT doctors can help patients purchase and set up remote equipment to measure biometrics, provide care, and talk face-to-face over the internet i.e. telemedicine. Doctors then are able to receive the data they need to adequately modify care plans without requiring a patient to walk into the office as well as have more frequent communication and therefore a better understanding of a patient’s day-to-day health status. IoMT not only allows for better continuous care but also boosts patient satisfaction and engagement. Patients that spend more face time with their providers tend to have better relationships and therefore better patient satisfaction—a critical component of healthcare with more and more models shifting to value-based reimbursement from health payers.

IMoT implementation roadmap:

While the Internet of Medical Things has the potential to fuel HC growth, IMoT implementation sought to be a rocky path. But approaching IMOT implementations with a pragmatic approach leads to a better navigation path. Let us evaluate some basics steps of IMoT roadmap.

  • Identifying Healthcare organization business goals to build IMoR ecosystem
  • Develop a viable and convincing business case to roll-out IMoT
  • Next coming up with a clear vision and goals to realize with connecting medical devices
  • Big-Bang approach may lead to burn-out, and hence identify pilots or PoCs od IMoT success areas
  • Take an iterative approach to reiterate the ideation process and move forward with an implementation initiative

Sounds generic! That is the stepping stone for IMoT implementation. Imagine that healthcare companies manufacture more than half a million different types of medical devices, including wearable external medical devices like insulin pumps, blood glucose monitors, etc, implanted medical devices – implantable cardioverter defibrillator devices, and stationary medical devices – scanning machines, etc. to name a few. Most patient interactions with the HC system involve the use of medical equipment and devices. IMoT brings these interactions to life. Hence taking an incremental approach is the only way forward.

The true implementation of IMoT involves, “developing an in-depth understanding of end users”, “defining funding, business and operating models”, “clearly understand device interoperability requirements”, “embed security at the core”, “ensuring regulatory compliance”, “more importantly attract talent and build digital capabilities”, “improve the adoption of medical technology at scale and with trust”, and finally “create an ecosystem of seamless partnerships”.

IMoT Solution Providers:

Colleagues on this forum have highlighted many advantages of IMoT like cutting emergency room wait times, remote health monitoring, ensuring critical equipment availability, improved drug management, optimized staffing and workflow, better diagnoses, better outcomes with fewer false alarms, etc. As IMoT value proposition is gaining more traction, many solution providers are offering products and solution to tap this value.

With an estimated market value for IMoT technologies >$150 billion in over next 3 to 4 years, Philips, Siemens, GE Healthcare and Medtronic are currently leading IoMT technology investments, with Philips primarily dealing with cardiac monitoring, remote patient communication devices and sensor-related products, and GE and Medtronic instead focusing on cloud-based technologies in existing monitoring devices, implants, and cardiac pacemakers.. Listing below few examples.

  • IMoT and Telehealth: Health Net Connect offers various remote patient monitoring packages that monitor conditions like CHF, COPD, diabetes, and hypertension with devices like BP/BG monitors, Handheld ECGs, pulse oximeters and spirometers. Not only is this technology leading to reduced costs as patients handle everything in-house, but by eliminating the need to visit health professionals and vice versa, it’s also improving their overall patient experience.
  • IMoT and Drug Management: Proteus Discover is a health company that measures medication treatment effectiveness and helps physicians improve clinical outcomes and patients reach health goals through sensor-embedded pills like the one mentioned above. Once the ingestible sensor-containing pill reaches the stomach, it sends a signal to patch the patient is wearing, which monitors each time a pill is taken, as well as their general rest and activity patterns. another example is, Abilify MyCite approved by the U.S. by the Food and Drug Administration
  • IMoT and Medical Device Monitoring: e-Alert from Philips are also ensuring that critical hardware is always accessible, and if something like a breakdown does happen, staff members will be immediately alerted.
  • Siemens IoT solutions for the medical device industry are powered by combining big data with digital twins, a virtual representation of actual devices, moving in tandem across the lifecycle and connected by digital threads. By connecting virtual development and production planning environments with real support and lifecycle production data, Siemens equipping med-tech organizations with the transparency and advanced analytic tools required to gain a competitive edge using big data.
  • eVisit is a telemedicine platform that enables doctors to conduct examinations and prescribe remedies for their patients by remote.
  • Amiko.IO focuses on providing products for respiratory disease management, complete with an AI-powered platform.
  • InfoBionic’s MoMe Kardia provides remote monitoring of cardiac arrhythmia.
Challenges implementing Healthcare IoT / IMoT:

HC firms have to overcome a few key challenges ranging from data security to legacy infrastructure that may hinder health care IoT initiatives. Alongside these evident challenges, IMoT should address the following areas for widespread adoption.

  1. Health data explosion and sensitivities: HC is one the largest sector contributing to massive data creation. HC organizations to use IMoT technology effectively have to address growing data storage needs. As well HC has to be exceptionally careful to treat patient data from IoT devices according to federal and state regulations. The flood of data created by the IoT gadgets and devices used in the HC industry could also cause unforeseen problems if organizations are not equipped to handle it properly and verify its quality.
  2. Lack of EHR system integration. While the data that is collected from IMoT devices can include a patient’s vital signs, physical activity that information does not typically travel to an EHR system and, in most cases, is not centralized or made easily available to providers. This limits the information’s value since it is not always presented to the provider in a clinical context.
  3. An increase of available attack surfaces with IoT devices: IMoT devices explosion in health care present concerning vulnerabilities as device use rises, so does the number of ways hackers could infiltrate the system and mine for the most valuable data. Hackers could potentially learn about how a connected medical device operates by getting into the system and reading its error logs. The knowledge the hackers gain could facilitate breaking into a hospital network or making devices publish incorrect readings that influence patient care. It is high time for vendors, providers, and manufacturers’ to collaborate to reduce patient risks by closing the gaps that can form between the layers of an IMoT system by reinforcing standards and normalizing secure protocols. It’s not possible to know all the cybersecurity risks health organizations may face. Nonetheless, facilities planning to implement IoT technology must take care to increase awareness of existing threats and understand how to protect networks and gadgets from hackers’ efforts.
  4. IMoT data in silos due to interoperability challenges: Patients are likely to collect different sets of data when using different medical devices depending on each device’s purpose and, in some cases, the ordering physician. IMoT data alone may not be as meaningful if it is not within the context of a full health record. With the lack of wider adoption of adequate interoperability, data from different IMoT devices may remain locked in each individual system and lose its potential value to the rest of a patient’s care team.
  5. Data security causes concerns in the IMoT implementations: From the time that the data is collected at the device level to the point that it is transmitted over to its final destination, securing that information is critical and is required under HIPAA. But with the lack of common security standards and practices, many health IT professionals have concerns about the risks associated with IMoT device tampering and data breaches.
  6. Plan for ecosystem needs to be successful: According to a recent Cisco survey, ~60% of projects encounter trouble at the PoC stage or shortly thereafter. The study suggested that utilizing external partnerships (e.g. platforms) was a crucial factor for those organizations that achieved successful implementations. When it comes to the starting small and prioritizing projects that align with their most prominent business objectives or patient needs is key to the success.
  7. Overcoming legacy infrastructure challenges: Outdated infrastructure is a known fact in HC. Even though retrofitting can breathe new life into aging infrastructure, truly taking advantage of IoT is tricky if a facility’s infrastructure is outdated. Hence using IMoT in ways that make sense for the needs, budgets, and infrastructures of HC organization and having robust plan to ramping up resources to fill the gaps is the key to the success of IMoT implementations.
  8. Stringent high availability and near-zero tolerance for failure: One of the common use of IMoT technology in HC is to apply it to patient monitoring systems. While it is handy to take that approach, unlike other IT systems (ex: websites), these devices typically cannot go through planned periods of downtime. Hence, updates have to occur seamlessly as people use the monitoring devices. For the hospitals to depend on IMoT-enabled supply cabinets to track resources reducing inventory management issues, IMoT devices devices are to be audited correctly eliminating human errors.

AI in Operations (“AIOps”)

AIOps

Recently I was searching for verbatim “AIOps” on Google and got 624K results. Without many surprises noticed that there have been over 100 times rise in search trends since July 2017. That signifies the momentum for AI led Operations.

As my curiosity on AIOps increased, I looked at market opportunity for AIOps. From MARKETSandMARKETS analyst data, the global AIOps platform market size is expected to grow from USD 2.55 billion in 2018 to USD 11.02 billion by 2023, at a Compound Annual Growth Rate (CAGR) of 34.0% during the forecast period (2018–2023).

In this blog post, I am attempting to capture some highlights gathered from my learning curve over a past year or so. Refer to the schematic above that provides a high-level “AIOps Framework”. The following are key elements of the framework.

“AIOps” Verbatim Defined: Simply stating AIOps stands for Artificial Intelligence for IT Operations. Extending AIOps to business operations is inevitable in near future. Adding further, AIOps automates various aspects of IT and utilizes the power of artificial intelligence to create self-learning programs that help revolutionize IT services

AIOps Context: There is a significant opportunity to leverage AI for analyzing enormous data being created by IT and business operations tools, to increase the efficiency of operations, speed up services delivery and ultimately create superior user experiences. The resulting power of AIOps is enabling the progress from siloed to integrated operations backed by intelligent insights.

Signals: In today’s business and IT operations environment, the user is adapting multiple channels of communication for ease and enriched experience. So the backend operations teams as well should expand their ability to sense, analyze and respond to such structured, unstructured and semi-structured data signals. With this in mind, the AIOps platforms are being developed with built-in capabilities to receive and response signals that can encompass any events, alerts, service requests, IoT sensor data, Email, Video, Text, Voice support, UX, Social channels and many other forms.

Interfaces: The way enterprise operations backbone interfacing with signals and external queries also is shaping up in this transformation.

  • The first layer is Machine-First: Giving software/machine/bot the first act on sensing and responding to operations requisitions not only improves the automation of repetitive tasks but also augments cognitive intelligence in complementing human intelligence.
  • Human-Next Touchpoints: Human-next layers take up the operations requisitions that are not solvable by machines. These are the requests which involve human interventions.
  • Ensuring Reliability of Services: Alongside the above two layers, taking an engineering approach to services reliability for constant monitoring, triaging and incorporating insights from advanced analytics of enterprise data brings the culture of continuous improvements and stability to operations.

AIOps Platform: The entire AIOps ecosystem is based on the underlying Platform and Enterprise Core that ties all the components together. As Gartner defined, “Artificial Intelligence for IT operations (AIOps) platforms are software systems that combine big data and AI or machine learning  functionality to enhance and partially replace a broad range of IT operations processes and tasks, including availability and performance monitoring, event correlation and analysis, IT service management, and automation.”

As businesses are increasingly software-driven, operations downtime is becoming more costly and slow is the new down. This is leading businesses to proactively manage and improve experiences of services, applications, cloud, and networks. Along with this business 4.0 is digitally shifting the businesses offering the technologies that increase the volume, velocity, and variety of data. As traditional systems and manual efforts are facing challenges in correlating and analyzing the data or alerts, AIOps is stepping up to augment the enterprise intelligence in operations.

To conclude, the future is bright for IT and business operations with AIOps. The increasing shift of organizations core business toward the cloud, raising investments in the AIOps technology ecosystems, exponentially growing data volumes and increasing end-to-end business application assurance and uptime are driving the growth of AIOps market demand.

 

 

Predictive Maintenance Value for Process Industries

PM

Process industries are undergoing digital transformation building and integrating Minimum Viable Products (MVPs) in their strategic path to enabling business models, services, customer experience, operations, and workplaces re-imagination. What I notice across process industry segment is application of industrial internet concepts in creating predictive maintenance models that are yielding advantages including – greater machine availability, superior process quality, easier to plan service intervals, longer machine service life, safer and more sustainable operation, lower service efforts and decreased costs. I am highlighting few aspects demonstrating thought leadership in this space.

  • Companies are sponsoring proof of concepts and pilots for creating models to monetize predictive maintenance. As Predictive Maintenance and Condition Based Monitoring directly impact equipment uptime, by offering Predictive Maintenance as a service, the manufacturer can guarantee equipment uptime to their customers for a fees, i.e. selling value-added services which promise recurring revenue.
  • Process manufacturing is leveraging integrated utilities to reducing electricity consumption with just-in-time energy management with a dynamic platform delivering energy performance improvement with ‘as-a-service’ through edge connectivity of various assets, data acquisition and gateway, cloud-based technology, and analytics. Also include tracking people movement and asset utilization.
  • SRP performance monitoring center using Industrial Internet is another classic example. Since starting the GE Digital’s SmartSignal program in 2012 and through to 2016, SRP identified more than 1,900 issues, of which 800 were “catches” – a problem that the plant was not previously aware of and, with the new alerts, was able to take corrective action. With time and improved training of the algorithms, the rate at which the company identifies true issues and catches has improved.
  • One use case of specific interest to Food and Pharma industry’s glass packaging quality control and improvement is Wi-NEXT IIoT that drives major changes in glass container quality improvement reducing non-conforming products by 7%, which equals 5% extra line productivity, better process control, and higher customer satisfaction
  • Lastly sustainable business models of predictive maintenance includes – bundling within basic service agreement framework, a freemium offering during warranty with downstream revenue potential, offer value added service with pay-per-use model, and gain-sharing with partner ecosystem.

Process manufacturing winners are those who identify best in class practices for developing business models for predictive maintenance of equipment. Win-win scenarios for manufacturers arise from enabling collaboration of experts in this space to exchange ideas, spot trends and drive innovations.

Operational Excellence in New Age Businesses

operational-excellence

Operations is an integral part of businesses and operational excellence leads to business excellence. It is designed to provide value to customers and stay relevant with customer needs in achieving customer satisfaction. Excelling in operations is an art to ensure the success of new business models being innovated and and deliver substantial stakeholder value quarter on quarter in addition to expanding revenues on ongoing basis. Fundamentally the following are pillars of operational excellence.

  1. Product Design: Businesses have to design and develop products that are industry renowned and recognized by analyst communities to echo the sentiments. Managing delivery of minimum viable products designed is the default expectation. The underlying business finances create and reinforce the strong backbone for success.
  2. Process Optimization: A simple measure of process efficacy is delivery excellence. Right from business alignment to managing right infrastructure to ensure the customer experience is mandatory in digitally reimagined world.
  3. Support Efficiency: Customer focus and result certainty is the norm of business support. End user and customer always pose a healthy challenge and it should be primo motto of businesses across the world to reinforce the customer engagement and avoid customer dissatisfaction. The key to achieving this is building the right talent and demonstrate the workforce results.
  4. Operational Effectiveness: The time tested leadership and governance is the foundational block of operational excellence. Among all the industry renowned practices workforce communication and engagement plays a key role here. Fiscal and regulatory compliance is at most important along with Deming & Japanese quality measures. Corporate social responsibility and ethics is defacto expectation of millennials to name a generation importance.
  5. Art of Innovation: Innovation at least should demonstrate financial viability and sustenance of a business. If a business is not creating a constant value to society its very existence is questioned and rest is the story based on historical anecdotes. Innovation comes from promoting culture of joint investment from partners and create an ecosystem of co-innovation. Remember we are living in a world of crowdsourcing.
  6. Security & Safety: It has maximum relevance in today’s connected world. Security comes from assurance – the level of guarantee that a system will behave as expected, countermeasure – way to stop a threat from triggering a risk event, risk threat avoidance, vulnerability management and exploiting vulnerability that has been triggered by a threat. Safety is important to distinguish between products that meet standards, that are safe, and those that merely feel safe.

I will be discussing further details around tenants of operational excellence in next post.