Machine data analytics company Glassbeam has just raised $2m in funding (read here). Puneet Pandit, CEO of the Santa Clara, CA-based startup, answered our questions about the market, the product, its achievements to date, the funding round and future plans.
FinSMEs: Hi Puneet. First, can you tell us a little bit more about you? What’s your background?
Puneet: I am electrical engineer with a MBA in finance and marketing with more than25years of global IT experience. Prior to Glassbeam, I was founder and CEO of Orchesys, a professional services firm focused on enterprise storage solutions. Glassbeam was incubated inside Orchesys and launched to the market in 2009.Specifically, I am responsible for setting the company direction, hiring some of the smartest talent from the industry and delivering on the promise of ground breaking machine data analytics solution. Prior to Orchesys, I was senior director at Network Appliance – leading the Database and Business Applications Solutions Group, specifically focused on driving $300M “Oracle on NetApp” market with cross functional initiatives across joint R&D, sales, services and marketing programs. Before NetApp, I worked at Ernst & Young strategic advisory services and Tata Unisys as a management consultant.
FinSMEs: Let’s speak about Glassbeam. What’s the opportunity you found in the (IoT) market?
Puneet: At Glassbeam, we are laser focused on a mission to create a new category of IoT analytics solution around machine log data. The premise is that any high technology product (storage, server, switch, medical device, industrial machine, connected car, etc.) is increasingly capable of generating copious amounts of rich but unstructured machine data, that once mined, can provide tremendous business and operational insights for its manufacturers and users.
Broadly, IoT is a hot trend driven by three key drivers: Sensors,Connectivity and Analytics.
Sensors: People do not realize it, but sensors are everywhere -your car and phone are full of them. More complex devices like storagesystems and MRI machines also have hundreds of sensors generating allkinds of machine data every few minutes. There are three key reasons forthis growth – sensors continue to shrink in size; they are cheaper; andthey use less and less power. A good example is the MEMS accelerometers,which are being widely used in airbags in cars. The cost of these devicesquickly dropped from hundreds of dollars to tens of dollars. At the lowercost, they could be used in all types of other things. Now these devicesare used in billions of smartphones, and the price today is down to adollar. As sensors get smaller and cheaper, more markets open up. In theprocess, sensors are changing the world by adding smarts totraditional devices. So how do you capture these “smarts” from the sensor embedded devices? That is where connectivity comes into play.
Connectivity: We all know the power of wireless today. Just imagine thatthese connected ”things,” devices and complex systems that are generatinglog data, start transmitting this data as a stream to a central location (cloud) that capture all the intelligence embedded in this data (on health and status of these systems). We do not have to really imagine this – it is happening today. Many smart companies are collecting and sending thisdata on a regular frequency to a cloud-based platform for deeper analysis.And no one wants to delete this data manufacturers want to store themfor years so that they can mine it to better understand and predict devicefailures, or better understand usage behavior with their end users. Thereare plenty of interesting use cases once you have this massive data storedin one place. But how do you uncover this hidden value in thissemi-structured machine log data? That is the last leg of this perfect storm – the analytics.
Analytics: Never before in the IT history has it been possible to analyze billions of data points at a fraction of a second and at a fraction of the cost (all terms relative to past). Today, we have the power of opensource technologies – Hadoop, Cassandra, Mahout, etc. – to store all this data, parse it, analyze it and create extremely compelling business andoperational value of any company. You can spin up servers on Amazon at afraction of a cost – pay as you go, and scale up or scale out on the fly.You can deploy some remarkable open source statistical algorithms from RSystems or Mahout libraries on this data and predict device or partfailures. You can even deploy a team of data scientists from off shore locations, such as India, to play with the data and create business value unheard of before. All in all, analytics is the last mile where raw data gets converted into actionable intelligence.
And that is where Glassbeam enters into the play. Today, as an analytics company, we have an enviable technology and production-ready solution thatcan scale to terabytes of streaming machine data in the cloud. We havesome great examples from companies like IBM, Aruba Networks and Dimension Data that we have documented in case studies, and they will stand up and talk about their positive experiences with Glassbeam solutions in the IoT space. We are pushing now hard to go beyond our roots of IT systems to non-IT machines like medical devices where we can apply the same fundamentals of our core
technology and provide value to thesemanufacturers for improving support operations, building new revenues withvalue add services, and gathering product usage/intelligence to buildbetter products.
FinSMEs: Tell me something about the features of your solution…
Puneet: Glassbeam in an end-to-end cloud-based platform and application suite focused on providing value to any manufacturer of “connected” machines, devices or things (IoT). Our platform’s core features include a highly scalable, performant parsing and ETL engine that can ingest, process, organize and analyze machine log data of any format (variety), streams (velocity) and size (volume).
We offer standard features as a part of the platform:
– Multi tenancy: Cater to data from multiple manufacturers, products and end users, in one single architecture in the cloud.
– Security: Provide secure services in our hosted cloud offering based on current regulations, standards and best-practices, including compliance with a wide variety of standards such as SOC 1/SSAE 16/ISAE 3402 (formerly SAS70), organizations such as ISO, and federal programs such as HIPAA, CSA and FedRAMP.
– Scalability: Provide highest levels of scalability with modern NoSQL architecture of Cassandra, MapReduce, translating to horizontal and vertical scaling options to meet dynamic workloads from multiple customers at the same time.
For applications, we provide the following key features:
– Log management: Cloud based “Log Vault” that allows any authorized user to search and download raw logs for any machine or customer.
– Search and analyze: Glassbeam Explorer application provides a rich interface to search full-text or through facets. Searches can be saved, shared, and illustrated with graphical output.
– Ad-hoc analytics: Glassbeam Workbench allows user to play with parsed data and do “what-if” analysis, export data into xls, graph, chart and build dashboards.
– Proactive and Predictive analytics: Glassbeam Rules & Alerts app provides the ability to input rules, take action, open cases in CRM and send proactive email notifications.
– API and Web Services: Glassbeam Direct Access feature allows access to underlying parsed data using APIs and extracts (SQL access) to use third party apps like Tableau to chart, graph and build dashboards.
FinSMEs: What’s the real advantage it brings? Is it easy to use?
Puneet: Our core value prop is to reduce mean time to resolution (MTTR) for customer escalations in a support function (buyer is VP Support). Secondary value prop is to help product managers (buyer is VP Products) gain unprecedented insights into actual product/feature usage by mining “truthful” machined data (machines never lie). Finally, we help sales and service organizations (buyer is VP Field Ops) to help increase revenues by identifying up- sell and cross-sell opportunities by gaining proactive account intelligence.
Solution is very easy to use – just like any SaaS app through a web-based browser and user logins. Setup is fairly straightforward – a few days to couple of weeks to get the solution up and running. More customizations can be made as things progress with more active usage and discovery of key business needs with end users.
FinSMEs: Where are you now in terms of growth? Some numbers?
Puneet: We are about 35 employees, 5 major customers(IBM, HDS, Aruba Networks, Meru Networks, Dimension Data). We are tracking about 20,000 connected machines in our cloud today, a combination of storage and wireless devices from customers mentioned above. Collectively, we ingest and process about 7.5 billion sensor readings per day from this growing population of devices and customers. Our data warehouse, growing into terabytes, stores information over years for deeper reporting and analytics across all these customers.
FinSMEs: You just raised a round of funding. What can you tell me about the investors? How are you using the funds?
Puneet: We just raised $2M in new funding bringing total funding into the company to $8.1M. This was an Angel led round with lead investors from VKRM Group. Our plan is to use these funds to accelerate our sales pipeline, build a better positioning through specific marketing programs, and enhance our product portfolio for IoT analytics market.
Puneet: IoT market is evolving rapidly and there is a dearth of proven technologies and solution providers, specifically in the IoT analytics space. We feel fortunate to be part of this growth trend and are looking forward to expanding our footprint to become a leader in IoT analytics market.