At one of the many ‘server shut down ceremonies’ to take place at media monitoring firm Isentia over the last year, someone from the IT team pasted a picture of the retiring server onto a pi?ata. “Then the infrastructure and ops guys got to smash it!” remembers chief information officer Andrea Walsh. “We’ve had lots of different parties and ceremonies as they’ve turned off.” When Walsh joined Isentia in 2015, the Sydney-headquartered company which operates across the Asia Pacific region had been struggling with its on-premise set-up. SUBSCRIBE TO OUR NEWSLETTER From our editors straight to your inbox Get started by entering your email address below. Please enter a valid email address Subscribe The media’s attention is often unpredictable, meaning when focus fell on one of Isentia’s long list of clients, there were huge spikes in compute demand for its monitoring services. “Really early on we knew we had an environment that was constantly having issues and disruption to service,” Walsh told CIO Australia at the Amazon Web Services Re:Invent conference in Las Vegas last month. “We had a large physical footprint and a challenge with that is how do you cope with spikes in volume of data? It’s that tipping point where you need enough capacity to cope with a level of a spike, but then it sits there idle in between.” Two years ago, Walsh, with the backing of the executive, established a cloud-first strategy. Starting with its broadcast storage, then its client Mediaportal and social listening offering, the company is now – almost – all-in with cloud, with AWS. With only a few pinatas left to pop, Walsh and her team, freed from the fire-fighting and management, are looking at how far they can push automation and how machine learning might improve their customer offerings. “We can now focus more on creating new services and being a center of innovation for Isentia,” Walsh said. Public perception Isentia operates in 11 countries across the APAC region supporting more than 5,000 clients worldwide, including 84 of the world’s top 100 brands. Among them, the likes of IBM, Lenovo, L’Oreal, Macquarie, Microsoft, NAB Bank, ANZ Bank, HSBC, Nike, Nikon, PG, Samsung, SAP, Seek, Starbucks, Visa, Walt Disney, AMP, Bupa, Canon, Destination NSW, and Domain. The communications teams within each of those companies needs to know when and where their company is being mentioned and it what way. They might also want to keep track of how their industry or competitors are being discussed. The same is true of the public sector – the Australian Government is a major customer. “What we do is set up a series of search criteria all across TV, radio, online, social media, print press and basically find anything relevant for them,” Walsh explains. “It could be how your CEO is being represented, or how the AGM was perceived, in the public sector how is the public responding? How’s the media responding to this suggestion or recommendation? And it’s not always what I would assume to be obvious things – it’s often something that’s emerging that is completely separate from anything,” she adds. Isentia’s clients access coverage through the company’s Mediaportal, and can access a repository of media files about them. The volume of data monitored daily by Isentia is in Walsh’s words “insane” averaging about nine million items a day. Isentia captures for its clients every day: 13,000 broadcast summaries30,000 print items300,000 Facebook comments/posts1,200,000 online news items3,200,000 forum posts4,500,000 tweets daily Isentia FY17results presentation, August. “We try to find relevance in all the noise and there is a lot of it!” Walsh says. Page Break Walsh started in the cloud by migrating Isentia’s broadcast storage to Amazon S3. Next went Mediaportal and social-media-listening servicesand the development of a cloud-based data pipeline to support the services and integrate the company’s disparate systems The pipeline utilises Amazon EC2 for compute instances, Amazon RDS configured in a high-availability mode across multiple Availability Zones, Amazon Kinesisstreams to capture social-media conversations across the web andAWS Lambdato perform preliminary computations on data before it gets deposited inAmazon Elastic MapReduce. “You can have these really crazy explosions and spikes, moving to the cloud is really great for us because we could just scale. And it can cope with that,” Walsh says. New focus The move to cloud has transformed Walsh’s IT team. A restructure meant a number of staff were retrained and experts employed. The team is chiefly made up of engineers, plus those in infrastructure and operations. Walsh has improved capability with training in Agile, Scrum and Kanban; and introduced Lean and Design Thinking. There have been hackathons and a dedicated innovation lab has been established. Andrea Walsh, CIO Isentia “Our primary focus now is on building out features in Mediaportal because they’re going to benefit clients, and continuously building out this data pipeline that we have,” Walsh says. Isentia was given access to beta versions of three of AWS’ new products launched at Re:Invent: Amazon Transcribe (which can transcribe voice audio recordings into clean text), Translate (for translating text fromn one language to another – “We have clients that need thousands of translations a day in Asia because we’ve got 10 different languages there,” Walsh says) and Comprehend (which can pick out positive and negative sentiment in text). Each has helped Isentia move staff onto higher value tasks, Walsh says. “These services allow us to be able to use the technology to do the heavy grunt work and then use the people to refine their results,” she says. Image recognition tool Rekognition is also being tested to pick out logos in television broadcasts. “It’s okay when you have a really obvious clear object and there’s not a lot of noise around it. If you’re trying to pick up something maybe on somebody’s top or a sports person and their moving it’s really been very impossible to try and do, with very mixed results!” Walsh says. “We’ll be looking to integrate [the services] further into our pipeline so we can get more efficiencies and automation.” At last count the performance of Mediaportal has increased by 25 per cent and the performance for social-listening products has increased by 40 per cent since the cloud migration, Walsh says. Administration time has been reduced by at least 50 per cent and it’s also enabled new products like Stories. Stories analyses and clusters media coverage into stories in real time using algorithms capable of calculating the connection between tens of millions of data points. Isentia suffered a shaky 2017 – its share price dropped more than 50 per centover the 12 months – with a legal dispute with rival Meltwater and the closure of the content marketing firm King Content it paid out $48m for in 2015. Now focused on its core business, its cloud foundations mean it is back on a more steady footing. “We see FY2018 as a year of cementing our transition from a media-led to a data- and technology- led organisation,” Isentia’s departing chairman Doug Flynn said in the company’s last annual report. The transition has not been easy. “Somebody new comes in, they don’t have thousands of clients like us. They can build and go whoops let’s try something else – not a major impact to them,” Walsh says. “Certainly moving [to cloud] has been really challenging. We’re doing the lifting and shifting, we’re building new, and also servicing existing clients. We can’t disrupt that. People say – oh just lift and shift – it sounds easy – it’s a lot of work!” The author travelled to AWS Re:Invent 2017 in Las Vegas as a guest of Amazon Web Services. Related content brandpost Should finance organizations bank on Generative AI? Finance and banking organizations are looking at generative AI to support employees and customers across a range of text and numerically-based use cases. By Jay Limbasiya, Global AI, Analytics, & Data Management Business Development, Unstructured Data Solutions, Dell Technologies Sep 29, 2023 5 mins Artificial Intelligence brandpost Embrace the Generative AI revolution: a guide to integrating Generative AI into your operations The CTO of SAP shares his experiences and learnings to provide actionable insights on navigating the GenAI revolution. By Juergen Mueller Sep 29, 2023 4 mins Artificial Intelligence feature 10 most in-demand generative AI skills Gen AI is booming, and companies are scrambling to fill skills gaps by hiring freelancers to make the most of the technology. These are the 10 most sought-after generative AI skills on the market right now. By Sarah K. White Sep 29, 2023 8 mins Hiring Generative AI IT Skills feature Top 17 cloud cost management tools — and how to choose Cloud cost analysis tools help your organization keep on top of its overall cloud use and associated costs, which can add up rapidly. By Peter Wayner Sep 29, 2023 14 mins Cloud Management Cloud Computing Podcasts Videos Resources Events SUBSCRIBE TO OUR NEWSLETTER From our editors straight to your inbox Get started by entering your email address below. Please enter a valid email address Subscribe