Quantiphi Talks ‘Enormous’ Gen AI, Google Cloud, Databricks Momentum
Google Cloud and AWS all-star AI partner Quantiphi explains customers’ generative AI demand, Databricks’ acquisition of MosaicML and Google’s new Premier Partner badges.
Artificial intelligence channel superstar Quantiphi is seeing an “enormous explosion” of new customer opportunities in the red-hot generative AI market thanks to massive investments from the world’s largest cloud companies, including Amazon Web Services and Google Cloud.
“We started the business 10 years back with an AI-first mindset. We were early. It was painful,” said Hasan. “Then the market has slowly converged around what we were hoping and attempting to do. For us, now is the perfect moment to say, ‘We now have a team of about 4,000 people. And with the generative AI capabilities that we built, we can continue to scale and be a bigger force in the market.’”
The Marlborough, Mass.-based company won AWS’ prestigious ML/AI Partner of The Year award in 2022, while also receiving Google Cloud’s North America Breakthrough Partner of the Year award in 2021. With such top-notch AI expertise, Quantiphi knows where the generative AI market is heading.
Databricks’ Acquisition Of MosaicML
Last month data lakehouse specialist Databricks struck a deal to acquire generative AI platform company MosaicML for $1.3 billion. The goal is to provide customers with the ability to build, own and secure generative AI models with their own data.
“Databricks and MosaicML have an incredible opportunity to democratize AI and make the lakehouse the best place to build generative AI and LLMs [Large Language Models],” said Databricks CEO Ali Ghodsi last month.
Quantiphi is already leveraging both Databricks and MosaicML’s offerings and believes the merger could be a huge win for a key element of generative AI: open source.
“There will be a need for an open-source ecosystem where enterprises can take these AI models, fine-tune them and make them their own,” said Hasan. “Any development on the open-source front is something that we can be very excited about. So this combination of MosaicML with DataBricks is a great idea.”
In an interview with CRN, Quantiphi’s Hasan talked about the new Google Cloud Premier Partner badges the company plans to earn, the benefits of Databricks’ acquisition of MosaicML and what customers are demanding from generative AI today.
What are customers seeking from generative AI solutions right now?
What we’ve seen is a transition from customers saying, ‘I don’t know much about this, but I have a board-level or a CXO-level directive to do something with generative AI so I have to wrap my head around it’ to now, ‘These are the three problems that we are trying to solve. Can you help us build, deploy and life-cycle manage?’
In some cases, it’s gotten to the point where they’re now saying, ‘We are convinced that we have to be making an aggressive bet on this. Can you help us scale our Center of Excellence so that we can stand up these capabilities within our own organization?’
So in a very short period of time, people have gone from not knowing much to knowing what problems to solve and looking for expertise and help to solve for that. And now they’re looking for expertise and help to scale these types of capabilities.
We started the business 10 years back with an AI-first mindset. We were early. It was painful. Then the market has slowly converged around what we were hoping and attempting to do. For us, now it’s the perfect moment to say, ‘We now have a team of about 4,000 people. And with the generative AI capabilities that we built, we can continue to scale and be a bigger force in the market.’
Where does the generative AI market stand today, and what’s one big market trend?
There was a very turbulent period in early 2023 as everyone observed this new thing called a ChatGPT. Lots of people had very, let’s say, engaging experiences with what it could do. But also, lots of people left with a fairly unsettled feeling in terms of their conversations.
So it created a lot of concern, but it also sort of opened everyone’s eyes to what the potential of AI truly is from a generative AI sense. Then there was this dynamic of innovation from OpenAI and Microsoft, Google and other partners within the ecosystem.
Today, there’s a very credible counterweight that is coming from the standpoint of open-source models.
So now there’s this tension between frontier models, which are closed-source, very, very large versus open-source models that can get very close to what these frontier models are able to do.
So we are seeing just an enormous explosion of opportunities at the top of the funnel. We are doing a couple of meetings a day on this, but the intent of customers to do more with generative AI is growing every single day.
You said open source is a key evolving element of the generative AI market. Talk about how the Databricks acquisition of MosaicML could help.
We leverage a lot of what MosaicML has done. Their pretrained models come in different form factors— there’s a 7 million parameter model and the 30 billion parameter model.
MosaicML has essentially put everything out there as open source: the data on which model is trained, the source code for the training and the model base itself. So there are various form factors that we use.
The combination of MosaicML as part of the Databricks ecosystem I feel will give customers the ability to fine-tune their own models for their industry, teach the model language of their business and teach the model tasks that are relevant for their enterprises to solve.
This is a very good development because there are the frontier models, which are very large, very capable, but closed source—which means that you have limited capability to fine-tune it, to make it very precise, and to solve problems that you’re trying to solve. Because it doesn’t know the language of your business, and it doesn’t know tasks specific to your business.
Frontier models are very good for a whole variety of tasks like summarizing content, writing email, speeches and all that type of stuff. So there will be some use cases that these frontier models like Google’s PaLM, OpenAI’s GPT 4, etc. will solve very well.
But there will be a need for an open-source ecosystem where enterprises can take these AI models, fine-tune them and make them their own. Any development on the open-source front is something that we can be very excited about. So this combination of MosaicML with DataBricks is a great idea.
Google Cloud is one of your most strategic partners. Do you think you will achieve the new Google Cloud Premier Partner badges?
Yes. For Google Cloud, we are active across all three Premier Partner badge areas: Sell, Service and Build. We have actually achieved the Premier level in selling and service.
And because we have now started to build and deploy software and assets onto Google Cloud, and sell our own software and platform—we are also in the coming quarters going in for the Build premier badge level, which we should be able to achieve. That’s a stamp of approval that says, ‘This partner in this domain knows what they’re doing. And you can expect a high level of competence.’
How hard is it for a partner to achieve one of Google Cloud’s new Premier Partner badges?
There is a fairly extensive evaluation process that we have to go through to get the partner badge. At the Premier level, essentially, there is an external auditor who will be hired—it’s almost like a financial audit where you have to go through the number of customers that you serve and showcase individuals with certain certifications and years of experience and all that stuff.
So it’s a very deep, external third-party objective process that is done. Of course, because it has to be objective, there is a price tag on that for us to be able to get that independently as well. It’s a very thorough process.
What’s your view on Google Cloud’s new Premier Partner badging?
Google is receiving feedback from customers that there is a greater degree of expertise that is required. Right now, the tiering structure is such that it’s not as fine-grained as it needs to be on the expertise level.
The expectation at a certain level or a certain tier of partnership is that people come in with the technology expertise and domain expertise to serve customers in their environment This sort of creates another expertise level or credentialing where the customers can be certain to expect a certain level of competence. That’s essentially what’s happening with these Premier badges.
To us, it feels like just another evolution in the majority of Google’s partner programs.
What incentives are in store for Google Cloud partners who achieve a Premier Partner badge?
The incentives come in multiple forms. There are sort of incentives in place for mechanisms like a higher Delivery Readiness Index score. It’s also in the form of deal close. Then in a software sense it’s in the form of partner preference in certain areas for specific types of workloads and given industries, etc.
So there are soft benefits and hard benefits. But over time, we’ve seen that there is a very strong correlation in us moving up the tiering mechanism from a credentialing perspective and the growth and profitability of our business.