Dreamforce 2017 Recap
When we started SpringML, we knew it was vital to our growth that we develop strong relationships with leading cloud vendors. It was also clear who the first companies we should partner with would be:
- Salesforce – The launch of Salesforce Analytics required a new generation of partners to serve customers. Sales users were looking for dependable forecasting and pipeline analytics solutions, and we were convinced that SpringML had a role to play in this space.
- Google – Google’s cloud capabilities and a rich portfolio of big data and machine learning (ML) products made them popular with customers. We were interested in partnering with Google to better serve enterprise customers working with extremely large data sets (like video data) that require machine learning and analytics performance beyond the capabilities of the big data applications typically used in marketing and sales.
We quickly made inroads with Google and Salesforce, achieving several milestones with both partners, including:
- Being one of the first to build applications on Salesforce’s Einstein Analytics
- Publishing open source connectors for Salesforce to Apache Spark
- Publishing open source connectors for Salesforce using Google Dataflow
- Being a launch partner for Google Dataflow
Since then, our relationship with Salesforce has grown significantly, and we’ve completed over 100 implementation projects in data intensive industries such as energy, media, healthcare, and high-tech. And by partnering with Google, SpringML has been able to leverage our machine learning capabilities in applications beyond the industries and functions we’ve traditionally targeted, including semiconductor manufacturing and preventive maintenance for fleet management. At the intersection of these applications, we are starting to solve some pretty interesting applied machine learning use cases (for example, using the Einstein Intent API to classify cases or Google Cloud to predict mass failure in manufacturing).
We are very excited about the strategic partnership announced by Google and Salesforce last week. Through their relationship, Salesforce will more tightly integrate Google Analytics 360 and G Suite with the Salesforce core clouds, giving Salesforce users the ability to seamlessly connect sales, marketing, and advertising data for the first time. To get a better idea of just how impactful analytics can be in marketing efforts, take a look at our recent case study on machine learning and ad campaign management. The partnership between Google and Salesforce will play an important role in enabling the more precise management of ad campaigns in real time.
Looking to the future, we believe Salesforce and Google will continue to be heavily involved in machine learning and artificial intelligence. Both companies have made significant investments in this space.
The strength of the Salesforce platform is that it empowers analysts and everyday users to develop applications without requiring them to know how to code. We believe Salesforce’s tradition of developing user friendly platforms will make it easy for non-technical marketing teams to use Salesforce’s Einstein Analytics software, which should have the performance required to support the ML and AI needs of most marketing applications.
However, customers dealing with larger, more complex data sets need a framework that supports building an end-to-end pipeline and widely-used ML libraries. As nicely explained by Glenn Solomon of GGV, Multi-Cloud will be the new normal as clients choose platforms that provide the right level of resources. Adoption of TensorFlow and Google Cloud products such as PubSub, Dataflow and BigQuery will provide the solution framework for data scientists to solve complex use cases.
Of course, there is overlap between Google and Salesforce. We are often asked how the Einstein Language API differs from Google NLP APIs. This is bound to create confusion as customers try to determine which ML/AI offering is better suited for the specifics of their application. While we do see a potential for some minor overlap, we believe Salesforce and Google implementations of ML and AI will have different audiences. Let me explain by examining three aspects of the Salesforce and Google partnership we look forward to helping clients in 2018.
APIs and Integration
As we wrote in this article several months back, APIs will be a popular vehicle to expose specific ML functions. Clients are looking for expertise in integration of ML APIs to their applications. Given SpringML’s deep expertise in data integration, we are excited to solve complex integrations, any day.
Data scientist citizens
Extending the power of machine learning to non data scientists is a hard problem to solve, but it is becoming a business imperative. Salesforce newly released myEinstein empower data analysts and everyday users to apply machine learning on their data.
Supplementing human judgment with Machine Learning.
We strongly believe Artificial Intelligence without humans is stupid. At SpringML, we always put people and their work at the core. Success of any ML implementation depends on the interpretation of the results and accuracy of the models. We already seeing customers demand transparency of the result and explanations of the predictions instead of just a predicted value. Google Cloud provides the level of scale and transparency that allows us to build robust applications and services to serve our clients.
Working with Salesforce and Google teams for the last few years helped us to learn from the best and understand to approach user problems in completely different perspectives. Our vision remains unchanged and feel that the announced strategic relationship validates our chosen path.
We are very positive about 2018 and the years to come.