programming languages
Auto Added by WPeMatico
Bodo.ai secures $14M, aims to make Python better at handling large-scale data
Bodo.ai, a parallel compute platform for data workloads, is developing a compiler to make Python portable and efficient across multiple hardware platforms. It announced Wednesday a $14 million Series A funding round led by Dell Technologies Capital.
Python is one of the top programming languages used among artificial intelligence and machine learning developers and data scientists, but as Behzad Nasre, co-founder and CEO of Bodo.ai, points out, it is challenging to use when handling large-scale data.
Bodo.ai, headquartered in San Francisco, was founded in 2019 by Nasre and Ehsan Totoni, CTO, to make Python higher performing and production ready. Nasre, who had a long career at Intel before starting Bodo, met Totoni and learned about the project that he was working on to democratize machine learning and enable parallel learning for everyone. Parallelization is the only way to extend Moore’s Law, Nasre told TechCrunch.
Bodo does this via a compiler technology that automates the parallelization so that data and ML developers don’t have to use new libraries, APIs or rewrite Python into other programming languages or graphics processing unit code to achieve scalability. Its technology is being used to make data analytics tools in real time and is being used across industries like financial, telecommunications, retail and manufacturing.
“For the AI revolution to happen, developers have to be able to write code in simple Python, and that high-performance capability will open new doors,” Totoni said. “Right now, they rely on specialists to rewrite them, and that is not efficient.”
Joining Dell in the round were Uncorrelated Ventures, Fusion Fund and Candou Ventures. Including the new funding, Bodo has raised $14 million in total. The company went after Series A dollars after its product had matured and there was good traction with customers, prompting Bodo to want to scale quicker, Nasre said.
Nasre feels Dell Technologies Capital was “uniquely positioned to help us in terms of reserves and the role they play in the enterprise at large, which is to have the most effective salesforce in enterprise.”
Though he was already familiar with Nasre, Daniel Docter, managing director at Dell Technologies, heard about Bodo from a data scientist friend who told Docter that Bodo’s preliminary results “were amazing.”
Much of Dell’s investments are in the early-stage and in deep tech founders that understand the problem. Docter puts Totoni and Nasre in that category.
“Ehsan fits this perfectly, he has super deep technology knowledge and went out specifically to solve the problem,” he added. “Behzad, being from Intel, saw and lived with the problem, especially seeing Hadoop fail and Spark take its place.”
Meanwhile, with the new funding, Nasre intends to triple the size of the team and invest in R&D to build and scale the company. It will also be developing a marketing and sales team.
The company is now shifting from financing to customer- and revenue-focused as it aims to drive up adoption by the Python community.
“Our technology can translate simple code into the fast code that the experts will try,” Totoni said. “I joined Intel Labs to work on the problem, and we think we have the first solution that will democratize machine learning for developers and data scientists. Now, they have to hand over Python code to specialists who rewrite it for tools. Bodo is a new type of compiler technology that democratizes AI.”
Powered by WPeMatico
Okta launches a new free developer plan
At its Oktane21 conference, Okta, the popular authentication and identity platform, today announced a new — and free — developer edition that features fewer limitations and support for significantly more monthly active users than its current free plan.
The new ‘Okta Starter Developer Edition,’ as it’s called, allows developers to scale up to 15,000 monthly active users — up from only 1,000 on its existing free plan. In addition, the company is also launching enhanced documentation, a set of sample apps and new SDKs, which now cover languages and frameworks like Go, Java, JavaScript, Python, Vue.js, React Native and Spring Boot.
“Our overall philosophy isn’t, ‘we want to just provide […] a set of authentication and authorization services.’ The way we’re looking at this is, ‘hey, app developer, how do we provide you the foundation you need to get up and running quickly with authorization and authentication as one part of it,’ ” Diya Jolly, Okta’s chief product officer, told me. And she believes that Okta is in a unique position to do so, because it doesn’t only offer tools to manage authorization and access, but also systems for securing microservices and providing applications with access to privileged resources.
Image Credits: Okta
It’s also worth noting that, while the deal hasn’t closed yet, Okta’s intent to acquire Auth0 significantly extends its developer strategy, given Auth0’s developer-first approach.
As for the expanded free account, Jolly noted that the company found that developers wanted to be able to access more of the service’s features during their prototyping phases. That means the new free Developer Edition comes with support for multi-factor authentication, machine-to-machine tokens and B2B integrations, for example, in addition to expanded support for integrations into toolchains. As is so often the case with enterprise tools, the free edition doesn’t come with the usual enterprise support options and has lower rate limits than the paid plans.
Still, and Jolly acknowledged this, a small to medium-sized business may be able to build applications and take them into production based on this new free plan.
“15K [monthly active users] is is a lot, but if you look at our customer base, it’s about the right amount for the smaller business applications, the real SMBs, and that was the goal. In a developer motion, you want people to try out things and then upgrade. I think that’s the key. No developer is going to come and build with you if you don’t have a free offering that they can tinker around and play with.”
Image Credits: Okta
She noted that the company has spent a lot of time thinking about how to support developers through the application development lifecycle overall. That includes better CLI tools for developers who would rather bypass Okta’s web-based console, for example, and additional integrations with tools like Terraform, Kong and Heroku. “Today, [developers] have to stitch together identity and Okta into those experiences — or they use some other identity — we’ve pre-stitched all of this for them,” Jolly said.
The new Okta Starter Developer Edition, as well as the new documentation, sample applications and integrations, are now available at developer.okta.com.
Powered by WPeMatico
Pulumi brings support for more languages to its infrastructure-as-code platform
Seattle-based Pulumi has quickly made a name for itself as a modern platform that lets developers specify their infrastructure through writing code in their preferred programming language — and not YAML. With the launch of Pulumi 2.0, those languages now include JavaScript, TypeScript, Go and .NET, in addition to its original support for Python. It’s also now extending its reach beyond its core infrastructure features to include deeper support for policy enforcement, testing and more.
As the company also today announced, it now has over 10,000 users and more than 100 paying customers. With that, it’s seeing a 10x increase in its year-over-year annual run rate, though without knowing the exact numbers, it’s obviously hard to know what exactly to make of that number. Current customers include the likes of Cockroach Labs, Mercedes-Benz and Tableau .
When the company first launched, its messaging was very much around containers and serverless. But as Pulumi founder and CEO Joe Duffy told me, today the company is often directly engaging with infrastructure teams that are building the platforms for the engineers in their respective companies.
As for Pulumi 2.0, Duffy says that “this is really taking the original Pulumi vision of infrastructure as code — using your favorite language — and augmenting it with what we’re calling superpowers.” That includes expanding the product’s overall capabilities from infrastructure provisioning to the adjacent problem spaces. That includes continuous delivery, but also policy-as-code. This extends the original Pulumi vision beyond just infrastructure but now also lets developers encapsulate their various infrastructure policies as code, as well.
Another area is testing. Because Pulumi allows developers to use “real” programming languages, they can also use the same testing techniques they are used to from the application development world to test the code they use to build their underlying infrastructure and catch mistakes before they go into production. And with all of that, developers can also use all of the usual tools they use to write code for defining the infrastructure that this code will then run on.
“The underlying philosophy is taking our heritage of using the best of what we know and love about programming languages — and really applying that to the entire spectrum of challenges people face when it comes to cloud infrastructure, from development to infrastructure teams to security engineers, really helping the entire organization be more productive working together,” said Duffy. “I think that’s the key: moving from infrastructure provisioning to something that works for the whole organization.”
Duffy also highlighted that many of the company’s larger enterprise users are relying on Pulumi to encode their own internal architectures as code and then roll them out across the company.
“We still embrace what makes each of the clouds special. AWS, Azure, Google Cloud and Kubernetes,” Duffy said. “We’re not trying to be a PaaS that abstracts over all. We’re just helping to be the consistent workflow across the entire team to help people adopt the modern approaches.”
Powered by WPeMatico
Google Cloud launches Game Servers, a managed cloud backend for games
Google Cloud today announced the beta launch of Game Servers, a managed service that provides game developers with the usual backend services for running their games, including multi-player games, in the company’s cloud. It’s worth stressing that these are not game streaming servers but solely meant to make it easier for game developers to build, scale and manage the backend services for their games.
The service sits on top of the Agones open-source game server, a project Google and Ubisoft first announced in 2018, and the Kubernetes container orchestration platform. As Google Cloud product manager Scott Van Woudenberg told me, the team is also reusing some parts of Anthos, Google’s service for managing multi-cloud Kubernetes clusters. And while Game Servers can currently only run on the Google Kubernetes Engine, the plan is to allow for hybrid and multi-cloud support later this year.
Quite a few gaming companies have already built their own on-premises server fleets, so just like in the enterprise, having hybrid-cloud capabilities is a must-have for a tool like this. Google will also make it easy for developers who already use Agones outside of Game Servers today to bring those servers into the same managed Game Servers ecosystem by registering them with the Game Servers API.
As Van Woudenberg noted, virtually every game now needs some kind of cloud backend, be that for multi-player features, match-making or keeping persistent game stats, for example. That’s true for indie developers and major game studios. Game Servers, ideally, will make it easier for these companies to scale their clusters up and down as needed. Game Servers also provides for A/B testing and canary tests, and, in future updates, it will include integrations with the Open Match matchmaking framework.
To get started, developers still have to containerize their game servers. For those companies that already use Agones, that’s a pretty straightforward exercise, Van Woudenberg said. Others, though, need a bit more help with that, and Google is working with partners to walk them through this.
Powered by WPeMatico
Google launches TensorFlow Enterprise with long-term support and managed services
Google open-sourced its TensorFlow machine learning framework back in 2015 and it quickly became one of the most popular platforms of its kind. Enterprises that wanted to use it, however, had to either work with third parties or do it themselves. To help these companies — and capture some of this lucrative market itself — Google is launching TensorFlow Enterprise, which includes hands-on, enterprise-grade support and optimized managed services on Google Cloud.
One of the most important features of TensorFlow Enterprise is that it will offer long-term support. For some versions of the framework, Google will offer patches for up to three years. For what looks to be an additional fee, Google will also offer to companies that are building AI models engineering assistance from its Google Cloud and TensorFlow teams.
All of this, of course, is deeply integrated with Google’s own cloud services. “Because Google created and open-sourced TensorFlow, Google Cloud is uniquely positioned to offer support and insights directly from the TensorFlow team itself,” the company writes in today’s announcement. “Combined with our deep expertise in AI and machine learning, this makes TensorFlow Enterprise the best way to run TensorFlow.”
Google also includes Deep Learning VMs and Deep Learning Containers to make getting started with TensorFlow easier, and the company has optimized the enterprise version for Nvidia GPUs and Google’s own Cloud TPUs.
Today’s launch is yet another example of Google Cloud’s focus on enterprises, a move the company accelerated when it hired Thomas Kurian to run the Cloud businesses. After years of mostly ignoring the enterprise, the company is now clearly looking at what enterprises are struggling with and how it can adapt its products for them.
Powered by WPeMatico
Google Docs gets an API for task automation
Google today announced the general availability of a new API for Google Docs that will allow developers to automate many of the tasks that users typically do manually in the company’s online office suite. The API has been in developer preview since last April’s Google Cloud Next 2018 and is now available to all developers.
As Google notes, the REST API was designed to help developers build workflow automation services for their users, build content management services and create documents in bulk. Using the API, developers can also set up processes that manipulate documents after the fact to update them, and the API also features the ability to insert, delete, move, merge and format text, insert inline images and work with lists, among other things.
The canonical use case here is invoicing, where you need to regularly create similar documents with ever-changing order numbers and line items based on information from third-party systems (or maybe even just a Google Sheet). Google also notes that the API’s import/export abilities allow you to use Docs for internal content management systems.
Some of the companies that built solutions based on the new API during the preview period include Zapier, Netflix, Mailchimp and Final Draft. Zapier integrated the Docs API into its own workflow automation tool to help its users create offer letters based on a template, for example, while Netflix used it to build an internal tool that helps its engineers gather data and automate its documentation workflow.
Powered by WPeMatico