We’re excited to announce that Databricks was named one of many 2024 Fortune Greatest Workplaces in Know-how™. This award displays our dedication to creating an surroundings the place Bricksters can thrive and innovate whereas serving to knowledge groups remedy the world’s hardest issues. Rooted in our origins on the UC Berkeley AMP Analysis Lab, Databricks is devoted to cultivating an open and clear tradition, driving the subsequent technology of knowledge intelligence, and shaping the way forward for knowledge and AI.
The Greatest Workplaces in Know-how record is very aggressive. Nice Place To Work, the worldwide authority on office tradition, conducts America’s largest ongoing annual workforce examine, based mostly on over 1.3 million survey responses and knowledge from corporations representing greater than 8.2 million workers this yr alone.
Don’t simply take their phrase for it. Hear straight from the Bricksters who make Databricks a Greatest Office in Know-how.
A Tradition of Information-Pushed Engineering
Shoumik Palkar, Sr. Employees Software program Engineer
I work at Databricks right now for a similar causes I joined the corporate almost 5 years in the past: the breadth of expertise throughout the engineering group, the breakneck tempo of innovation and studying, and the deep worth positioned on utilizing data-driven choices to make our product higher for our prospects. The three are intrinsically linked, with the fountainhead being the energy of the engineering crew. Our engineers—who convey many years of expertise from the world’s largest knowledge techniques of their previous lives—function in an area the place “cutting-edge” evolves week-over-week and steady studying turns into a necessity, whether or not it’s by maintaining with the most recent tutorial convention proceedings or via cross-team collaborations to seek out novel options.
Nonetheless, instinct and studying usually are not sufficient. On the middle of every resolution at Databricks is the info itself. Information is the good equalizer, offering a unified language that enables us to push for innovation throughout the product with confidence. These substances make Databricks an extremely rewarding place to work, each by fostering private development via mentorship by world-class consultants and by enabling people in any respect ranges to ship buyer worth by letting the info determine.
Analysis-Pushed Tasks
Jonathan Ellithorpe, Employees Software program Engineer
Databricks’ first-principles method to engineering empowers me to dig deep and remedy issues at their core. On my crew we construct the basic distributed techniques “constructing blocks” that assist companies to function effectively and reliably at scale, together with high-performance storage, distributed caching, low-latency cache invalidation, and dynamic sharding and cargo balancing. We additionally work intently with groups throughout the corporate to get deep into their code and architectures to uncover widespread patterns and establish the place new abstractions are wanted. Due to Databricks’ data-driven and extremely collaborative tradition, it’s simple for us to take a research-style method to our work the place we favor early experimentation and measurement to check hypotheses and leverage prototyping and a build-then-iterate course of to innovate shortly.
In addition to addressing challenges in our infrastructure utilizing a first-principles mindset, steady studying {and professional} development can be a key a part of my expertise at Databricks. With the corporate’s tutorial roots and dedication to publishing often, I’m continually inspired not solely to broaden my data and expertise but in addition to be a part of the bigger tutorial group. For instance, Databricks has sponsored my attendance at OSDI annually since I joined, which has been important for me to remain linked to the group and hold abreast of the most recent within the area, together with in essential new areas such because the intersection of ML and techniques. Publishing can be extremely inspired, and this yr we submitted a place paper on high-performance in-memory stateful companies, showing at HPTS 2024.
For my part, the mixture of a first-principles-based method to engineering, a data-driven tradition, and an emphasis on steady private development and improvement implies that at Databricks, there’s actually no restrict to what you are able to do.
Distinctive and Vibrant Tradition
Tao Tao, Director, Engineering
Databricks stands out with its distinctive energy of bringing collectively the very best of the trade, the open supply group, and tutorial analysis. We created a complete new customary of “good engineering” that’s visionary, elegant, and sensible. At our core, our inclusive and collaborative tradition fosters a seamless trade of concepts and experience throughout groups and disciplines.
This cross-team synergy is a core driver of our innovation. My current instance of this spirit is the event of our newest Lakehouse Collaboration product providing: Databricks Clear Rooms – a privacy-safe resolution for safe knowledge collaboration powered by Delta Sharing. This breakthrough emerged from the shut teamwork among the many trade consultants, partnership crew, belief and security specialists, and a number of engineering groups throughout the stack at Databricks. By harnessing numerous views and specialised data, the crew was in a position to design an answer that not solely addresses the present advanced challenges however can be future-proofing for upcoming AI workloads.
For many who search to hitch a dynamic and inclusive crew alongside world-class consultants and to make tangible impacts on the info and AI trade, Databricks presents an distinctive alternative.
A Accountable Method to AI
Jasmine Collins, Analysis Scientist, Mosaic AI
At Databricks, we prioritize the creation of protected and accountable AI techniques that finest swimsuit buyer wants. A current instance of this dedication is the ImageAI mannequin, collectively educated with Shutterstock for producing photos tailor-made to enterprise use circumstances.
In contrast to most present picture technology fashions educated on arbitrary web-scraped photos and captions, we took a extra accountable and punctiliously curated method. By coaching our mannequin from scratch on Shutterstock’s proprietary and vetted picture repository, the mannequin’s outputs usually are not solely commercially viable however may be trusted for a variety of purposes. By utilizing a well-governed dataset, we be sure that the mannequin’s capabilities and limitations are clearly understood, making it extra appropriate for its supposed use circumstances in enterprise environments.
This method highlights our broader dedication to accountable AI improvement throughout all of our initiatives. By prioritizing knowledge high quality, investigating dataset bias, and making certain transparency, we goal to create protected and dependable AI techniques.
Construct the way forward for open supply AI-driven knowledge intelligence. Be part of us!
Be taught extra about profession alternatives at Databricks and see present job openings.