Friday, 7 June 2019

Security Analytics and Logging: Supercharging FirePower with Stealthwatch

When we consider network threat detection, most of us immediately think of signature and rule-based intrusion detection and prevention systems (IDPSs). However, it is a little discussed fact that the very first intrusion detection systems, built back in the ‘80s, were actually based on anomaly detection!

Cisco Study Materials, Cisco Tutorials and Materials, Cisco Learning, Cisco Guides

Those pioneers understood that with the presence of zero-days and the lack of exhaustive black-lists, we needed to use the full range of analytical techniques at our disposal to be effective.

Those anomaly detection roots may not be so evident in today’s IDPSs however they were not totally lost. In fact, a whole new branch of network threat detection systems were developed that used those very same anomaly detection techniques. That heritage manifests itself, today in so-called Network Traffic Analysis (NTA) tools.

While IDPSs have made detecting the initial intrusion in the packet stream their relentless focus, NTA systems take a very different approach. They generally work on metadata generated from the network, often called network flows, so they can expand our scope of analysis in both time and space to become essential in post-breach analysis, incident response, and even threat hunting situations.

Well Cisco has arranged a family reunion!

We are proud to announce the combination of our best-in-class IDPS and NTA products, Cisco Firepower and Cisco Stealthwatch. The Security Analytics and Logging (SAL) solution brings the best of perimeter-based protection and detection with the power of visibility and security analytics over the entire network. We believe we have created the most comprehensive network-centric threat protection, detection, and response solution – something that only Cisco is in the position to achieve.

Raising the bar on Network Security


It is very well understood how IDPSs are effective in security protection: blocking activity that can be identified as a threat or violates some policy. However, we accept that threats still get through and that is why IDPS have robust rules-based detection engines based on content-inspection.

But what do we do with all these detections? What if the traffic cannot be inspected? What if decryption is not an option? What if the threat is spreading internally?

Security Analytics and Logging service is specifically designed to augment your Cisco Firepower deployment with security analytics, from the Stealthwatch Cloud platform, to drive improved threat detections and provide the insight needed for more effective protection.

It All Starts with Visibility


The foundation of the solution is the aggregation of the connection and detection logs from Cisco Firepower with the network flows that the Stealthwatch platform collects. Just think about that. A dataset that gives us unprecedented visibility into the entire breadth of your network from perimeter to access, from campus to branch. But that’s not all! That “general ledger” not only contains all the header-based metadata, but now also includes all the metadata and inferences derived from all the deep content-inspection the Cisco Firepower engine provides.

Now you might be thinking to yourself, “there are plenty of tools I can use to gain this type of visibility.” However, in practice the sheer volume, velocity, and variety of the data can lead to staggering costs. The Stealthwatch team has made working at these scales our speciality and because our back ends are optimally engineered for the security outcomes we desire, we can offer this visibility in a much more cost-effective manner.

Security Analytics Driving Rapid Response


With all that visibility comes the opportunity to apply security analytics that can detect breaches that have bypassed the content-inspection based rules at the perimeter.

The security analytics powered by Stealthwatch can achieve this by baselining normal behavior of endpoints on the network in a process we call entity modelling. These models are then used to detect malicious activity based on any changes in behavior and indicators of compromise. The Stealthwatch engine can then combine these observations with others that may come from other parts of the network or even the detection engine in Cisco Firepower to create reliable and useful alerts.

Through this, you get detection of internal and external threats based on the analysis of network telemetry and IDPS logs, all from within Cisco Defense Orchestrator (CDO) and from that same interface, you can modify your network-wide policy to immediately deploy a remediation strategy. In addition, CDO is fully integrated with Cisco Threat Response which allows you to build incident casebooks and drive response actions across the whole of the Cisco security portfolio.

Closing the Loop: Improving Protection through Policy Tuning


Up until now, I have discussed the during and after phases of an attack but with SAL we can close the loop and reason more effectively about the before phase. In this phase we, as security practitioners, try to understand what is actually on our networks and what activity is to be allowed or blocked.

We express this intent through policies that enshrine both threat defense and compliance considerations. But designing and managing these policies across an increasingly complex digital business has historically been a major challenge and can leave many organizations vulnerable to attack.

The insight that it brings to the game drastically improves the way you can make policy decisions from within CDO. Through this capability you can query the logs collected from Cisco Firepower devices to play out what-if scenarios and validate the correct behavior of the policy at the enforcement point. In addition, the extended visibility of the rest of the network that the Stealthwatch platform provides can even allow you to determine if traffic is bypassing your enforcement points.

You can then turn around and deploy these highly tuned policies across the entire portfolio of security products right from within CDO! This is an entirely new paradigm that is required to not only scale with your growing network but also help you seamlessly manage policies across your environment powered by intelligence and insight.

Through this, you get detection of internal and external threats based on the analysis of network telemetry and IDPS logs all from within Cisco Defense Orchestrator (CDO) and from that same interface you can modify your network-wide policy to immediately deploy a remediation strategy.

Thursday, 6 June 2019

20 Years of Wireless with the Wi-Fi Alliance

In 1999, the idea of connecting to a network wirelessly was mostly a dream. The only device one might want to connect was a laptop, and they were generally expensive and often restricted to the executive suite in larger enterprises. But 1999 was also the year that the IEEE 802.11 Working Group approved the IEEE 802.11b standard, the technological base of Wi-Fi.

Cisco Wi-Fi Alliance, Cisco Certifications, Cisco Learning, Cisco Guides, Cisco Tutorials and Materials

However, the mere existence of a standard written by a bunch of smart engineers is rarely sufficient to ignite a revolution. Wireless technology needed a savvy champion, an industry organization to market the technology to the world and ensure it really works in the hands of users. Aironet (acquired by Cisco in 1999) was one of six companies to recognise this need and co-founded the Wireless Ethernet Compatibility Alliance (WECA), also in 1999.

The WECA took on two vital tasks immediately: to ensure the technology really worked in a multi-vendor environment, and to find a better name, one that would resonate around the world. It succeeded in both.

In 2000, WECA changed the name of its technology from Wireless Ethernet to Wi-Fi. Today, the Wi-Fi label is recognized across the globe and Wi-Fi is so valued that, according to at least one survey, people would rather give up beer or their morning coffee than their Wi-Fi.

Technical diligence


In 2000, the newly renamed Wi-Fi Alliance issued its first interoperability certification for an IEEE 802.11b product, operating at 11Mb/s. This certification was the first of over 45,000 product certifications over the next twenty years. As Wi-Fi technology has expanded, Wi-Fi Alliance certifications have kept users confident that their devices will work with products from other vendors. In 2019, users can be sure of interoperable, reliable and secure Wi-Fi access using the latest IEEE 802.11ax standard (now branded Wi-Fi 6) at rates of up to 5 Gb/s.

In the early years of the Wi-Fi Alliance, it was not always clear that Wi-Fi was going to become the dominant wireless access technology. It certainly was not pervasive. I recall IEEE 802.11 Working Group meetings in 2001, where most engineers designing standards for the next generation of Wi-Fi did not even have Wi-Fi access on their laptops; during Working Group meetings we often had to borrow Wi-Fi cards from a big box at the front of the room.

Today, the idea of a laptop not having perfectly-working Wi-Fi connectivity built-in is alien. Every laptop has Wi-Fi, as does just about any device that generates or uses data. Over 30 billion devices have been made with Wi-Fi, from security cameras in homes to badge readers in enterprises to entertainment systems in cars, industrial sensors, and, of course, mobile phones. There are so many devices using Wi-Fi that by 2022, Cisco’s Visual Networking Index forecasts more than half of all global IP traffic will access the network using Wi-Fi. Unfortunately, this traffic includes my Wi-Fi enabled bathroom scale, telling the cloud each morning that I really should do more exercise.

Challenges along the way


Wi-Fi is not perfect and never will be, but the Wi-Fi Alliance has provided a forum for ongoing development and improvement. For example, a flaw in Wi-Fi security was revealed in 2001 in the form of the WEP Debacle, in which it was shown WEP actually provided very poor security. It was almost a death sentence, because Wi-Fi without security is close to useless. Fortunately, the whole Wi-Fi ecosystem, led by the Wi-Fi Alliance, quickly pulled together and defined WPA (as a temporary solution) and then WPA2 (as a solution that has lasted more than 15 years) to ensure Wi-Fi had appropriate security to meet users’ needs. Of course, you can never take your eye off the ball with security. The Wi-Fi Alliance has continued to promote improvement, most recently with the release of WPA3 (with significant leadership from my Cisco colleague, Stephen Orr).

The Wi-Fi Alliance does not always get it right in its certification programs either, but every experience improves the process, and some “failures” hold the keys to future success. The Wi-Fi Direct certification for peer-to-peer communications was technically successful, in terms of the number of certifications, but the technology didn’t see widespread use. The Wi-Fi Alliance has not given up on peer-to-peer communications, though. Instead, it has learned from the experience;  there are great hopes that the recently introduced Wi-Fi Aware certification will better meet user’s needs.

The WiGig program for 60GHz access is another example where the Wi-Fi Alliance continues to persevere. This activity started in the Wi-Fi Alliance back in 2010. WiGig is still not yet successful, but it represents a significant opportunity for new spectrum and new use cases. The Wi-Fi Alliance’s ongoing work and perseverance means it is an opportunity that still has an excellent chance of being fulfilled in the near future.

Despite the Wi-Fi Alliance’s “learning experiences” over the years, the key point is that Wi-Fi has been successful because it has always fulfilled a promise to enable anyone, anytime, anyplace to construct a cost effective solution to solve real user’s problems. And the problems Wi-Fi solves are evolving. In 2000, the problem was connecting a laptop. Today, it is connecting anything to everything in homes, enterprises, factories, transport and public spaces.

The key to fulfilling this promise has been the Wi-Fi Alliance members’ cooperation across the Wi-Fi ecosystem. The Alliance is a forum for making sense of the alphabet soup of standards from the IEEE 802.11 Working Group, and for developing additional specifications as necessary. It’s also the primary forum for bringing vendors together to ensure interoperability of basic Wi-Fi technology as it continues to develop.

Proud to lend a hand


Cisco is proud to have played a role in the Wi-Fi Alliance since 1999. The company has been a driving force in the Wi-Fi Alliance from the very beginning, as a Sponsor member influencing its strategic direction and as a participant in Task Groups and Interoperability Test Beds. The Wi-Fi Alliance has a provided a basic interoperable Wi-Fi platform for Cisco to provide innovative features that meet the particular needs of our customers; features including Cisco Compatible eXtensions (CCX), controllers with coordinated Access Points, Cisco CleanAir® interference detection and mitigation, location based solutions such as Cisco DNA Spaces, Application Visibility & Control, Hyperlocation, Flexible Radio Assignment (FRA) of dual 5 GHz radios, Software Defined Access, and Intelligent Capture and real-time telemetry. In many cases, Cisco has contributed our proven features back into the Wi-Fi ecosystem, ultimately with certification by the Wi-Fi Alliance.

After twenty years, the global economic value of Wi-Fi is almost $2 trillion per annum (as of 2018). However, it is not the only globally-used wireless data network. Many claim that cellular data, in particular 5G, will take over from Wi-Fi in several key market segments. But Cisco don’t see this as a game with only one winner.

Cisco project that both Wi-Fi and 5G will succeed, and in fact strengthen each other’s success. Wi-Fi will continue to grow to meet the needs of the local area (in unlicensed spectrum), and 5G will meet the needs of outdoor, high speed needs (mostly in licensed spectrum). They will be better together – especially if users can move between the systems smoothly.

To help bring that vision to life, Cisco recently introduced OpenRoaming, building on the Wi-Fi Alliance’s Passpoint certification, which will allow users easy and secure access to Wi-Fi networks globally via a cloud-based federation of access networks and identity providers – including mobile carriers.

I am proud to have been personally involved with the Wi-Fi Alliance since 2003, most of that time on the Board of Directors, including as Chair of the Board from 2006 to 2011. I participated in its 10 year and 15 year anniversary celebrations, and now its 20 year anniversary. I look forward to watching Wi-Fi continue grow and develop in the future under the guidance of the Wi-Fi Alliance.

Tuesday, 4 June 2019

Driving Simplicity and Convenience for our Customers and Partners

At Cisco, one of our guiding principles is simplicity and convenience for our customers and partners. We believe that seeking and ingraining feedback in the future design and roadmap is a key factor which enables us to continually improve our products and solutions to solve real customer issues. In that vein, we received important customer feedback in three critical areas.  Here is what you told us:

1. You want to see all of your purchases in a single view. Without full visibility into what you own and what you are using, your organization could fall prey to significant legal, financial and operational issues. Legal issues like software compliance and audits.  Financial issues such as over or under purchasing or ineffective contract negotiations. And operational issues such as poor utilization of hardware and entitlements or expired service and support contracts.  It is difficult, if not impossible, to properly manage what you can’t see.

2. You need to be able to easily view and control who has access to service transactions and data. IT Administrators need to be able see and manage who has access to what, at any given time. Roles change, people move in and out of an organization, projects start and stop. Admins need instant access and control to generate or re-host licenses, manage user roles, and be able to quickly turn off access to critical network assets and entitlements when needed.

3. There are too many tools and processes, along with multiple, uncoordinated touch points.  Network infrastructures are getting more and more complex every day. With more tools, more portals, more subscriptions, more services, you need a solution that will consolidate all of the touch points and connect the dots for you.


Your feedback drove a new solution



My Cisco Entitlements (MCE) is a new, secure, user-friendly solution to manage assets and entitlements including technical support, software upgrades and downloads – all in one place, on one platform. MCE provides complete end-to-end IT infrastructure transparency. Building on the power of Cisco Smart Accounts, it brings visibility and control together on one platform that provides access to all Cisco services, subscriptions, licenses, and devices throughout their lifecycle.

No more portal hopping


With MCE you can now view everything in one place, instead of many. Real-time insights provide a forward view into products and services along with activation and utilization metrics.

A streamlined dashboard provides a customized view based on pre-selected filter choices. You can instantly obtain status on your systems and equipment, location of components, asset warranty, expiration dates, and more.

The flexible MCE platform provides the ability to:

◈ Filter, sort, export, tag and organize assets and entitlements
◈ Assign assets to Smart Accounts/Virtual Accounts
◈ Open a new support case
◈ Request software version upgrades on the fly

Cisco Certifications, Cisco Guides, Cisco Learning, Cisco Tutorials and Materials

Providing insight into critical IT questions


IT managers and network administrators are confronted daily with questions that directly impact their organization’s investments. Questions like:

◈ Are we fully optimizing the utilization of our existing assets and entitlements?
◈ Do we need additional or fewer services, subscriptions, licenses, or devices?
◈ What is nearing expiration or approaching end of support?

MCE provides valuable and actionable insights and answers to these important questions. For instance, MCE can proactively identify what’s at risk and the changes required to optimize an organization’s investment to its maximum potential. Dashboards and filters show usage metrics as well as service and support contracts that are near expiration.  Additionally, an organization’s investments are protected with secure and consolidated user access management using MCE.

We’re not done yet.


In the future, MCE will offer self-service MACD (Moves, Adds, Changes, and Deletes) on assets and entitlements. It will be the unified entry point to access all of your Cisco products and services entitlements such as rehosting licenses, requesting an RMA, and registering products and services. Features such as device management APIs, customizable and actionable notifications and alerts, and legacy licensing capabilities will all be standard.

While this is a giant leap in the right direction, we are not done yet. We will continually strive to build upon the platform and deliver more value, insights and capabilities for our users. We appreciate the partnership and the straight talk with our customers and partners, which has enabled us to bring together this unique platform.

MCE delivers on our simplicity and convenience for customers and partners ethos and I look forward to sharing more in the future.

Saturday, 1 June 2019

MUD is officially approved by IETF as an Internet Standard, and Cisco is launching MUD1.0 to protect your IoT devices

With over 8 billion “things” being connected today, IoT security has undoubtedly evolved from a mysterious buzzword to one of the biggest real threats to our network today. According to Gartner, over 51% of survey respondents believe that cybersecurity is the number one technology-related challenge for IoT deployment.

Overwhelmed by the countless number of IoT security comments and stories, let’s try to demystify this seemingly complex concept. To begin, let me ask you three simple questions: What types of IoT devices are connected to your network? What behaviors are appropriate for these IoT devices? Is there an industry standard to follow while connecting these IoT devices? If you don’t know the answers to these questions yet, that’s when we say the IoT security risks are probably right around the corner staring at you.

What is MUD?


To answer the above three questions, Cisco has been working on a solution known as Manufacturer Usage Description (MUD) to arm IoT security with you.

The key idea of MUD is to facilitate device visibility and segmentation by allowing your network administrators to effortlessly identify the type of IoT device and define the corresponding appropriate behaviors for that device. To do this accurately, we are introducing a participant to the conversation: the manufacturer. IoT manufacturers are able to disclose to us what their devices are, and what network policies they need for the devices to correctly function.  This whitelist statement is something that customers can use to deploy access policies in their own networks without any guesswork.

As shown in Figure 1, an IoT device first sends out a pre-embedded MUD-URL to the network devices (e.g. switch & AAA server), through which the MUD-URL will be received by the MUD controller (software). According to the specific MUD-URL, a matching MUD file will be provided from the MUD file server and translated into policy format through the MUD controller, to then enforce the access control list to the device.

Cisco Study Materials, Cisco Certifications, Cisco Learning, Cisco Guides

Clear benefits to both customers and device manufacturers brought by MUD


If you get the overall idea of MUD so far, you may see that IoT device manufacturers and customers are two key stakeholders in the MUD ecosystem. MUD offers distinct benefits for customers and manufactures:

Benefits to customers:

◈ Automate IoT device type identification thus reducing operational costs

◈ Simplify and scale IoT device access management by automating policy enforcement process

◈ Reduce threat surface of exploding number of IoT devices by regulating traffic and thus avoiding lateral infections

◈ Secure enterprise network through standard-based approach

Benefits to manufacturers:

◈ Improve customer satisfaction and adoption due to reduced operational costs and security risks

◈ Enhance device security through standard-based onboarding procedure

◈ Differentiate device offerings with embedded network-based device security feature

◈ Reduce product support costs to customers by following an easy-to-implement process

In addition to these benefits, we’ve received positive feedback from our partners:

“MUD technology is valuable for Innovative Lighting. MUD technology will enhance our commissioning process by identifying our devices on the network. Furthermore, MUD technology will provide the appropriate access control policy promoting a more secure system. We look forward to working more with Cisco and the MUD technology.”

-Harry Aller, CTO at Innovative Lighting

 “MUD was selected to protect Molex IoT solution against malicious parties. MUD is a relatively simple solution to implement at the device level, light on constrained IoT devices but takes advantage of strong network infrastructure including network switches and authorization server. Our goal to reduce exposure footprint and the overall solution allows us to provide a level of security to our customers that is scalable and flexible at the same time. The ability to whitelist specific devices in the field allows us to lock down the network but also to respond quickly to events that may take place post deployment.”

-Mo Alhroub, Manager of Software Engineering at Molex

MUD is approved as an Internet Standard and released as RFC8520 by IETF


I am delighted to announce that MUD has been officially approved as an Internet Standard by the Internet Engineering Task Force (IETF) and is now released as RFC 8520. Meanwhile, MUD is also part of the NIST Mitigating IoT-Based DDoS project, and an optional component of the Open Connectivity Foundation’s framework now.

MUD 1.0 is ready


Besides the IETF approval, I am also thrilled to announce that we are launching MUD1.0, the first phase of the entire MUD solution. While MUD itself is an open standard, Cisco is pioneering our unique version by leveraging Cisco switch and ISE (Identity Service Engine, a AAA server) as the network devices shown in Figure 1.

In this Cisco MUD1.0 release, we focus on providing device visibility by enabling the IoT device identification inside the enterprise network. As shown in Figure 2, the IoT device sends out the MUD-URL to the switch and then passes it to ISE. The administrators will see the device specific information on ISE UI including the device model, manufacturer, etc. Specifically, MUD1.0 supports profiling IoT devices, creating profiling policies dynamically, and automating the entire process of creating policies and Endpoint Identity Groups. Furthermore, administrators can leverage these profiling policies to create Authorization Policies and Profiles manually for securely on-boarding IoT devices.

Cisco Study Materials, Cisco Certifications, Cisco Learning, Cisco Guides

To make the sophisticated story simple, through MUD1.0, you would know exactly what devices are coming to your network the minute they are connected. Even more, you can define policies for these IoT devices. Isn’t that amazing?!

With MUD1.0 released, future releases will more fully automate the policy control part. On top of MUD1.0 when ISE receives the MUD-URL to extract the visibility information, the MUD-URL will be passed to the MUD controller (software) which will then go out to the MUD server to get the MUD file and translate the content into policy (as shown in Figure 1). The network devices will then enforce the appropriate policy onto the devices. The whole process will be fully automated. Want more flexibility as well? No worries, we’ve got you covered! Before the automation process, you get the choice to edit the recommended policy as needed.

Friday, 31 May 2019

Data Evangelism: Oxymoron, Fluff, or Business Driver?

At first pass, data evangelism may sound more like an oxymoron than a corporate function. Most of us (and our dictionaries) associate evangelism with faith, while data & analytics is core to the scientific method. Evangelism is predominantly qualitative while data & analytics is the definition of quantitative.

In practice, data evangelism has become synonymous with spreading the good word of data.  Need to inspire your team to balance their gut-based approach to problem solving with data-driven insights? Call in a Data Evangelist.

However, if we delve beneath the surface, data + evangelism reveals a richer value proposition. Evangelism teaches us to practice what we preach. Lead by example. Be the change we want to see in the world. Data & analytics teaches us to measure what matters. Hypothesize, test, minimize our biases, refine, and always let our data be our guide.

-----------------------------------------------------------------------------------------------------------------------
If we marry the tenets of data + evangelism, the result is:  Practicing the data & analytical methods we preach. Leading others to leverage data as an asset via a data-driven approach. Challenge ourselves as data evangelists to be at the forefront of data-driven models and insights, especially in the most qualitative domains.
----------------------------------------------------------------------------------------------------------------------

Data Evangelism Needs a Model


In data science, once you understand the data and its significance to the business, the next step is to create, stress test and refine a model which presents a simplified version of the business problem or opportunity you’re seeking to address. This model is a first attempt to explain the workforce’s relationship to data and provide actionable insights into creating (or maintaining) a data-driven enterprise.

Cisco Certifications, Cisco Guides, Cisco Learning, Cisco Tutorials and Materials

The Axes:


◈ Data IQ — The level to which a person is capable of leveraging data & analytics relative on his or her role and goal. For example, a food coordinator who is data literate and comfortable using a simple forecasting model will have a high Data IQ. If, however, s/he wants to lead an engineering team responsible for a machine learning-based technology, a Master’s or PhD in AI will be the new standard for a high Data IQ.

◈ Data Enablement — The level to which a person is enabled (or unable) to leverage data & analytics relative to his or her role and goal. For example, a people manager in HR may be fully Data Enabled via: data literacy, foundational data science for leaders, a dashboard which provides him/her the relevant people analytics and insights about their team, access to data & analytical talent on a project-by-project basis, and a steady stream of curated content including training, best practice sharing, and success stories. However, someone managing a data science team would need all of that and much more, including tools and platforms which allow for reusable asset (i.e. models and code) sharing, to be Data Enabled.

The Quadrants:


◈ Enthusiasts — Low Data IQ; Data Enabled: Well connected to their data & analytics community, fluent in its success stories but unsure how to begin leveraging data. Example: A marketing new hire with a degree in literature who marvels at chatbots.

◈ Data Illiterate — Low Data IQ; Data Unable: Lack of understanding regarding the value of leveraging data & analytics as well as how to do so. Example: An experienced technical writer who leans into his/her qualitative strengths.

◈ Siloed High Performers — High Data IQ; Data Unable: Limited by their isolation. Typically start from scratch instead of having a library of assets at their fingertips and peers with whom to collaborate. Example: a data scientist working on a non-data science team without access to mentorship, peers, enterprise tools, platforms and data products/services.

◈ Data-Driven — High Data IQ; Data Enabled: Individuals have the platforms, infrastructure, tools, services, and knowledge to leverage data & analytics in their role. Connections into the larger community provide them with a constant stream of ideas, best practices, and opportunities to collaborate as well as share their work. This is the target state.

-----------------------------------------------------------------------------------------------------------------------
Data-driven workforces, whose employees have High Data IQs and are Data Enabled, power the most digitally disruptive companies in the world.

Should we start looking to data evangelism as a business driver?
-----------------------------------------------------------------------------------------------------------------------

Data-Driven by an Evangelism Engine


Cisco Certifications, Cisco Guides, Cisco Learning, Cisco Tutorials and Materials
How does this play out? Let’s say a Customer Success Executive leverages data that is 22% more accurate than previously possible to enable 96% adoption of the collaboration tools his/her customer purchased. The customer wins by realizing a high ROI; Because the customer wins, the Customer Success Executive wins. Evangelism’s “win” is in enabling the person or team behind the 22% increase in data accuracy and the Customer Success Executive to leverage said data to achieve (and know s/he achieved) 96% adoption.

Our Approach

Our efforts to influence Data IQs take the form of a multi-pronged (and evolving) strategy of recruiting, learning & development, and continuous education.

We approach Data Enablement more broadly. Success in this domain doesn’t just take a village, but rather the support of the entire Data & Analytics business unit in addition to strong cross-functional partnerships. Data Enablement encompasses building, buying, supporting and/or co-creating the data products and services needed to enable each role- as well as those products’ and services’ adoption.

While far from an exhaustive list, Data Enablement includes global virtual and live events, Kaggle-style data science competitions, collaboration platforms for technical and non-technical best (and worst) practice sharing, an enterprise data science platform with reusable asset libraries, and democratized trustworthy datasets… and as data & analytics (and data evangelism) matures, who knows?

Thursday, 30 May 2019

Cisco Demonstrates 20W+ Power Dissipation of QSFP-DD at OFC 2019

At OFC 2019 in San Diego, CA, Cisco demonstrated the thermal management capability of the QSFP-DD pluggable module form factor for 400G applications. The demonstration showed that QSFP-DD can dissipate the heat generated by a transceiver that draws more than 20 watts of power. This will be required for extended reach applications that use advanced optical modulation schemes such as coherent transmission. Host platforms with QSFP-DD ports can therefore support pluggable coherent modules needed for DCI (Data Center Interconnect) and WDM (Wavelength Division Multiplexed) networks.

This is a key point because data centers are often deployed in regions where they may be several tens of kilometers apart. To link them together, data center operators would typically use transponders, which connect to a switch or router and transport the data streams to another data center over relatively long distances with a coherent optical link. However, if coherent modules such as 400G-ZR and 400G-ZR+ modules could be plugged directly in to a switch or router, it would eliminate the need for the additional connection and space for a transponder.

With advances in optical integration and CMOS DSP technology, Cisco expects pluggable coherent modules that support the OIF 400G-ZR specification to begin entering the market in the next 12 to 18 months. These new transceivers may draw 15-20W. By demonstrating 20W capability, we show that Cisco 400G QSFP-DD based platforms will be coherent-ready when those modules are available.

Cisco Certifications, Cisco Guides, Cisco Learning, Cisco Tutorials and Materials
Figure 1: QSFP-DD >20W Demo at OFC 2019 with Nexus 3432D-S Ethernet Switch and Thermal Modules

The demo included the recently announced Cisco Nexus 3432D-S ethernet switch. This Nexus 3k switch is a 1RU fixed-port switch with 32 QSFP-DD ports providing data centers with industry-leading performance-per-watt power efficiency at low latency, offering leading analytics.

This switch provides an extreme case environment from a power dissipation perspective, which makes it ideal for demonstration purposes. It was loaded with 8 thermal modules dissipating a total of 174W clustered together in adjacent ports.

The thermal modules were provided by MultiLane and were specifically designed for this demo. These are not functional transceivers, but rather QSFP-DD mechanical housings with heat sources inside. Each module has a series of heaters distribute across its top and bottom, positioned to emulate the thermal profile of real modules with the temperature measurement point over the hot spot of the module, as typically specified.

Cisco Certifications, Cisco Guides, Cisco Learning, Cisco Tutorials and Materials
Figure 2: QSFP-DD Module with External Heatsink inserted in to Cage

To accommodate more than 20W, the modules were modified with an additional heatsink on the portion of the module that extends beyond the front panel. This additional material resides within the current envelope of the Type 2 MSA module dimensions of QSFP-DD (see Figure 2).

The chart below shows a plot of the eight modules in the demo and how the hottest module might be expected to perform over the operating temperature range of the Nexus 3432 switch. The demonstration showed that the 8 modules in adjacent ports were dissipating more than 21W each.

Cisco Certifications, Cisco Guides, Cisco Learning, Cisco Tutorials and Materials

At the show floor booth’s ambient temperature of 21C, the case temperatures of the modules ranged from 33C to 44C. This is shown by the colored dots at the 21C vertical axis, and corresponds to a rise in case temperature ranging from 12C to 23C  and averaging 18.6C. This means that if the switch were operating in its maximum operating temperature environment of 40C, the case temperature of the hottest module would rise to 63C (rightmost end of the green line). That provides plenty of margin for the typical commercial maximum case temperature of 70C.

In summary, this demonstration shows that the QSFP-DD form factor is easily capable of managing high power dissipation required for coherent transmission in platforms that will be shipping later this year. From copper to coherent, along with backward compatibility with prior generation of QSFP modules, QSFP-DD provides customers maximum flexibility to address the migration to the next step in networking.

Tuesday, 28 May 2019

Demystifying Artificial Intelligence’s Role in Contact Centers

AI is shaping the future of customer experiences and the contact center.

Artificial Intelligence, Cisco Study Materials, Cisco Guides, Cisco Learning, Cisco Certifications

Artificial Intelligence (AI) is creating a lot of excitement and there are good reasons for this. According to Forbes, 50% of IT professionals believe artificial intelligence and machine learning are playing a role in cloud computing adoption, growing to 67% by 2020. This week I will be speaking at UC Expo about the role that artificial intelligence is playing in advancing contact center productivity and efficiency, and in particular how this is benefiting the agent and customer experience.

With all the news around AI, it’s easy to get lost in the hype versus reality. I’d like to demystify some of this and share my view of the five most common myths I’m hearing about as I travel the world talking to colleagues, partners, and customers.

Five myths:

◈ AI is new
◈ AI = chatbots
◈ AI can replace all your people
◈ AI is all about automation
◈ AI will reduce call volume

1. AI is new


The concept of Artificial Intelligence (AI) has actually been around for a long time. In 1950, the English mathematician and computer scientist, Alan Turing documented his ideas for testing a thinking machine. Turing’s test theory suggested that if a machine was able to communicate in full conversation via a teleprinter without any detectable differences from a human, the machine could be deemed “thinking.”

Contact centers have been using AI in some form or another for decades. We just didn’t call it AI. Even with basic automatic call distribution (ACD) technology it was possible to filter and route calls to the right agent at the right time using an algorithm to determine the best agent. AI at its core is a series of fast predictions, and contact centers have been using predictive algorithms since the early 1990s. AI continues to show up in new places and make a major impact in revolutionizing contact centers and customer service.

2. AI = chat bots


Chat bots have become mainstream in contact centers and are just one of the ways in which AI is used to optimize agent workload and enhance customer self-service. Bots are being used for everything from qualifying customer requests, to booking hotels, and providing shopping assistance. Bots helps answer questions and direct your customers to the appropriate person with the best skills and experience. Our Customer Virtual Assistant provides a highly effective way of offloading simple, mundane, and repetitive requests from your agents, allowing them to spend higher quality time helping customers on more complex requests. This results in providing faster self-service to customers, while improving agent optimization, productivity and costs. While chat bots utilize AI, there are many other forms that AI takes in contact center operations beyond chat bots – including routing schemes, forecasting, deeply inspecting customer interactions – and more. Cisco’s Customer Journey Analyzer, for example is a cloud based solution that allows companies to gather data from multiple sources to draw correlations between operational and business data to improve customer experience.

3. AI can replace all your people


This is probably the most misunderstood belief about AI and one that grabs the attention of senior executives when they’re looking for ways to cut costs. AI promises benefits far beyond just cost savings, and has the potential to help and improve how your employees work, rather than replace them. AI will certainly change workloads, staffing and processes that may lead to reduced staffing, but a primary advantage is that it augments agents to make them more scalable and efficient. Take the example of a chat bot. When the bot detects that the interaction needs to be escalated to a human agent, it brings along with it the history of the conversation to enable a seamless transition from self-service chat to assisted chat. The agent can then very quickly and effortlessly take over the interaction with everything they need displayed right in front of them. In this regards, AI is an enabler for better live agent assistance, not a replacement of it.

4. AI is all about automation


Business process automation is a key benefit of AI in the contact center because it increases agent efficiencies and workflows, however it is better to think of it in terms of agent augmentation as much as automation. Many people still want to engage with a human, so as you take advantage of automation in your contact center, make sure your customers can still reach a person when one is needed. According to a study by PwC, only 3% of U.S. consumers want their experiences to be completely automated. In addition, automation of customer processes inevitably leads to exceptions that cannot be addressed by AI, and need human assistance. These exceptions are often make-or-break moments in a customer relationship.

The automated process should learn from the data that you have spread across a multitude of systems, and from human interactions so that those experiences are improved resulting in better service to your customers. Understanding and analyzing that data can tell you so much about how your customers are experiencing your brand, so that actions can be taken to make their journey better. A great example of this is our new Cisco Answers intelligent agent which is powered by Google Contact Center AI. Cisco Answers listens to customers conversations in real-time, and proactively presents intelligent suggestions, documents, and other key forms of enterprise knowledge to the agent desktop, while the agent is interacting with a customer. This empowers agents with the needed context and information they need to deliver faster, more personalized and proactive care.

5. AI will reduce call volume


This is another misconception – AI may reduce call volumes via automation – but it’s likely that the calls that do make it to live agent assistance (the exceptions) will be longer and more complex in nature. This could mean that the same amount of staff are handling less, but longer calls. AI enables more accurate decisions and routing so that many customer interactions can be handled without a human agent. It has the ability to classify information and make predictions faster and at higher volumes than humans can accomplish on their own. As an example, in an omni-channel contact center, customers can be routed much more quickly and efficiently to the right resource that can add the most value to the experience based on the channel (e.g. email, phone, chat) the customer is using. According to PwC, 46% of all consumers will abandon a brand if the employees are not knowledgeable. This will require contact centers to redefine how they do agent staffing and how they measure call volume and call success.

So where do we go from here?


While AI for contact centers isn’t necessarily “new”, advancements in algorithms and the ability to apply them in real-time to massive amounts of data being created from contact center operations is. This combination opens up exciting new possibilities for companies to break free of constraints to address issues that have been challenging contact centers for decades. These include contextual routing real-time work to agent attributes, more precisely forecasting agent schedules, and higher degrees of customer personalization. As more and more data accumulates, processor speeds increase unabated, and algorithms march forward, AI will have a larger and larger role in making contact centers more efficient and effective.