Jackson Baker

Trying out Databricks AI/BI Genie and wish you had some best practices to work with? The Genie team heard you loud and clear, and wrote up this doc to help! https://lnkd.in/gvQhPFrV As always, feedback welcome. :)

190

4 Comments

Fellow in-house taxonomists, ontologists, metadata managers, and semantic wizards: what is the name of the team or business unit that you work within? I have the good fortune of getting to rename the "Content Standards" program I lead at Kaiser Permanente IT. My team has determined that the name is too vague to convey the core activities of the program, and leads to confusion with the cross-functional teams we collaborate with. I'm looking for a clear name that will resonate with stakeholders within my org. So I've got a very classic taxonomist question: What to call the program? 🤔 For some context... What my program is responsible for: - Defining and governing standards for metadata, structured data, taxonomy, and ontology for classifying/structuring digital content - Implementing processes and platforms to good metadata hygiene - Assessment and remediation of content metadata What my program doesn't cover: - Writing or editorial style guides - Content design - Content strategy - Site maps or navigational structure - Data governance Any ideas are most welcome!

37

48 Comments

I wrote some thoughts and observations about Google laying off their Python support team. You can read them if you are so inclined. Blog post here: https://lnkd.in/gMFhuZ76

1

Great article by Roi Yarden Building Data pipelines comes with its set of challenges. One of the most important decisions is to understand what patterns suit the business evolving needs and allow for a system to scale. This can prevent very costly mistakes.

8

Some takeaways for a Jeff Hinton interview with timeless insights. My POV as an app builder is that models are now being improved for better UX (faster response/multi-modal) and reasoning capability and cost continues to improve. The technology headroom remains strong, and the integration of LLM into all software experience will continue and expansion of the reach (invasion) of software + intelligence will also continue. - Next word prediction is not merely parroting of training data. To do next word prediction well, LLM needs and already exhibits strong reasoning behavior. This would continue to scale / improve with more data / training. - LLM is a compression of human knowledge. The compression mechanism is identifying similar concepts across domains and reduces duplications among them. This can also be a source of "creativity" as LLM "sees" correlations of seemingly unrelated ideas through this compression. LLM can be more creative as it sees similiarity / correlations across all domains beyond (human) convetional wisdom. - LLM can exceed the quality of its training data. Jeff gave an example of an experiment where a model trained on 50% intentionally incorrectly labeled data, can still perform at 90+% accuracy after training. - Multi-modal models can improve the overall world understanding and lead to better overall reasoning. It does not even need to generate multi-modal output. - Not ML specific, but an interesting life insights. Bad intuition comes from believing (or wanting to believe) everything you are told. It diminishes the ability to make a clear choice that would lead to differentiated success (or failure).

4

I've seen a few people in my feed calling out Andy Jassy for not mentioning or giving any credit to the employees of Amazon for their recent financial success (in their public filings). While that's accurate, I don't know that's it's fair to criticize Jassy for this, at least in comparison to the rest of his tenure as the CEO there. In particular, there's no data to support the idea that Jassy cares about the employees at all; remember, this is the same person who hired people as remote employees during the pandemic, and then brutally shifted the company's remote work policy in direct opposition to both the data and employee feedback, and effectively fired anyone who would or could not comply (which in some cases, meant upending their lives and families to relocate for the company). Jassy is a relentless cost-cutter, who (to his credit) is relentless focused on _his_ customers, the shareholders of Amazon. I'm 100% certain that Jassy would eliminate 99% of the employees of Amazon tomorrow, without any second thoughts, if he thought it would raise the net profits by even 5%. I'm certain Jassy views the human employees of Amazon as just an unfortunate business expense, and only marginally contributory to his personal efforts to optimize the company's profit margins. There are some really good/smart people who work hard to keep Amazon successful, and overall they have done a pretty good job as a company, and this post isn't intended to take anything away from that. I'm just noting that Jassy demonstrating that he doesn't care about, or even acknowledge, the work of the rest of the employees there is 100% in line with who he appears to be as a person and CEO, in my experience.

1

Amazon-to-Google-to-Amazon #a2g2a thought of the day: interesting to see how RTO (Return-to-Office) outcomes are entirely different for the two companies. In my post today, I don't really want to go down the path of debating whether RTO is "good" or "bad". There's very emotional and valid opinions on both sides, and it affects people's personal lives in real ways. Please be respectful. Whether you agree with the policy or not, for it to actually work it does need *critical mass*. Both companies have essentially the same RTO policy (3 days/week expectation), but there's significant differences is in how the message was delivered and enforced - and the outcome. Google’s 3-days-per-week was mostly a “pretty-please, would-you, if-you-don't-mind” with no actual enforcement (but some vague threats). So most people ignored it. They still do. Leadership got frustrated and the message got less polite over time, but nothing really changed. Even as of 3 weeks ago, when I was still a Google employee, the buildings were fairly empty. That made being physically in the office significantly less useful, other than enjoying the perks of free food and massages by myself (and occasionally with 1 or 2 coworkers). Amazon’s 3-days-per-week came with actual enforcement. Amazon used badge reader data to track when people were actually coming to the office, with consequences. It does feel a bit draconian, but you can’t argue with the effectiveness of saying “we have a policy that we believe is business-critical and we’ll enforce it or else.” I appreciate that when I come to the office, I can find my coworkers there and we can have f2f meetings. There’s critical mass for RTO to actually be effective. Amazon has picked a lane. Personally, I have not figured out how to truly be productive and happy working remotely in the last 4 years. I respect and appreciate the fact that some people have, but I haven’t. I do like the option to stay home a day or two, skipping my long commute, working from a lounge chair in my backyard on a sunny day. But there’s nothing like the energy in the room of a good group meeting, or the chemistry of a face-to-face 1:1, or the enthusiasm of two engineers standing in front of a physical whiteboard drawing boxes and arrows with markers brainstorming about the future. Or the bonding of sharing a meal or a walk between buildings with a coworker. I feel energized when I’m physically in the office. It literally charges my batteries. I’ve been coming to the office 4x/week and loving it. 

183

34 Comments

Large Language Models (#LLMs) are evolving rapidly, and the demand for running them at scale is skyrocketing. #Ollama is stepping up to meet this need with its official container image on Docker Hub, excelling in local #deployments. However, there’s a catch for containerized environments: downloading #models requires the ollama serve command to be running within the #container. This extra step complicates automated deployments via #CI #CD #pipelines, making seamless model access challenging. To take your #model to production, it needs to be #container-ready—a feature Ollama has yet to fully support. The gap in #AIOps for smooth deployments gets blurred with the inference services available as an easy way to take it to production. Goal is to overcome obstacles and deliver solutions 😉 so here is to get you started on having your (O)#llama #production ready in a private #deployment from your #local to #cloud all in #oneshot . Enjoy Medium: https://lnkd.in/g9jqBQ8d dev.to: https://lnkd.in/gi_gGwam #AI #LLM #Ollama #llama3 #phi3 #mistral #S6overlay #AIOps #CI #CD #oneshot #container #docker #podmanEnjoy

11

There is a broad phenomenon in tech companies right now (based on anecdotal LinkedIn research) of culture shifts. I want to give you a model for thinking about this problem. The problem can be thought of as an "explore/exploit" problem, specifically one with limited time and resources. Tolerance for exploration is correlated with more time and resources. Exploration in company cultures includes things like investing in long term learning programs, trying new benefits, testing theories, and generally being more flexible with planning and experimentation. Appetite for Exploitation increases when there are options with well established upsides compared to options with high uncertainty, and resources are constrained. These constraints can be real, imagined, or manufactured. This is a special form of risk aversion. Problematically, humans are bad at the math for knowing what the optimal stopping point is for either exploitation or exploration, and often pick the wrong strategy for the moment. But the trend right now is exploitation, both culturally and economically. (Not to be confused with abusive exploitation, though that may happen in step with certainty exploitation.) I'm curious - does this describe a situation you've seen?

2

3 Comments

���� Attention Engineering Managers and Technical Leaders! 🌟 In just two weeks, on June 28th at 10 AM PDT, I’m excited to host an interactive roundtable on Feature Flags and Delivery in conjunction with ELC! 🚀 Why Attend? This is a collaborative discussion where your expertise and curiosity are key! Whether you're experienced or eager to learn, your contributions are invaluable. 🔍 Discussion Points: - Decoupling Processes - Boosting Team Efficiency - Optimizing Feature Delivery - Tracking Platform Usage Expect rich conversations on best practices, unique implementations, methodologies, and the latest tools enhancing feature flagging and delivery. Who Should Join? - Engineering Managers - Technical Leads - DevOps Engineers - Software Developers This roundtable is your chance to: - Share your experiences - Gain new insights - Network with peers - Discover practical recommendations RSVP Now and mark your calendars for June 28th at 10 AM PDT! Let’s collaborate and elevate our skills together! #EngineeringManagement #TechLeaders #FeatureFlags #DevOps #SoftwareDevelopment #TeamEfficiency #TechRoundtable

9

New blog post: The little things that got me into stream processing (https://lnkd.in/gU7Ba5Ym)

60

To all of my UltiPeeps that were impacted by the layoffs at UKG today, please don’t hesitate to reach out if you need a recommendation on LinkedIn, need help preparing for your next opportunity, or just want an old friend to talk to. Stay strong in this environment, and know that there is life beyond UKG. Sometimes the greatest opportunities in life present themselves in moments like these. #UltimateSoftware #UKG #Layoffs

149

2 Comments

Navigating the Pitfalls of Rapid Scaling: A Lesson in Cultural Cohesion Targeting CTOs, VPs of Engineering, and CEOs, this discussion highlights a common challenge during post-Series C expansion: integrating top-tier talent without sacrificing cultural cohesion. In a past experience, our team aggressively hired industry-leading engineers to boost technical prowess and market standing. However, emphasizing technical skills over cultural fit led to a fragmented team dynamic, impacting project timelines, innovation, and morale. Here are actionable insights for leaders navigating similar challenges: Strategic Onboarding: Beyond technical orientation, emphasize immersion into company culture, mission, and values. Cultural Fit Assessments: Include cultural alignment in hiring criteria to ensure new hires complement existing team dynamics. Transparent Communication: Regularly communicate the company’s vision, helping new hires understand their role in broader goals. Leadership Development: Encourage leadership skills that align with your company ethos through mentoring and peer learning. Feedback Mechanisms: Use feedback loops for employees to express concerns and suggestions about integration. Balancing rapid hiring with cultural integration is key. By valuing human elements as much as technical ones, we create a scalable, innovative, and harmonious engineering environment. I invite fellow leaders to share their strategies for maintaining cultural cohesion during rapid growth. What has worked for your organization?

5

I keep reading this every couple days and falling in love with it more. We keep having more and more systems because people need to write articles and books. And, of course, you have to have your own acronym. But when we step back and look at the core of things, most of it boils down to: 1. Go to therapy and deal with your own shit 2. Talk to other people. Be humble. 3. Actively listen to them and think deeply about what you heard from them 4. Don't be a dick. Have empathy. 5. Try to figure out situations where everyone wins That's it.

6

1 Comment

Interesting comment in speaker feedback from my Monitorama talk last week: > Anecdotally our most impressively traced applications are our ML/AI/LLM applications. This talk made me appreciate that team more and what we can learn from them that applies to other teams. This is something I've observed when talking with several folks in orgs (ML engineers, SREs, product owners) -- the fact that "works on my machine" is such a poor proxy for how an ML system will behave when your users hit it forces teams to adopt better instrumentation, ground their system understanding in real data, and use that data to inform how they improve things. The crossover between ML and "ops" concerns with Observability is huge.

23

1 Comment

Looking to get started with in-database machine learning with MySQL HeatWave? This guide will help get you on the right path. https://lnkd.in/gPjuJU9g

1

Harvard Business Review recommendations for getting hired: #4 Tailor your Resume I recently stumbled upon a super useful tool - EmberLime (https://emberlime.com/). I especially loved their UX (code review-esque style) It effortlessly customizes your resume in minutes to maximize ATS matching score to be in the Top applicants faster than anyone.

13

1 Comment

While the corporate culture and return-to-office policies differ "markedly" between the three companies, the similar effects of the RTO mandates suggest that "the effects are driven by common underlying dynamics," wrote the authors of the study. "Our findings suggest that RTO mandates cost the company more than previously thought," said David Van Dijcke, a researcher at the University of Michigan who worked on the study. "These attrition rates aren't just something that can be managed away." Robert Ployhart, a professor of business administration and management at the University of South Carolina, said executives haven't provided much evidence that RTO mandates actually benefit their workforces. "The people sitting at the apex may not like the way they feel the organization is being run, but if they're not bringing data to that point of view, it's really hard to argue why people should be coming back to the workplace more frequently," Ployhart said. Senior employees, he said, are "the caretakers of a company's culture," and having to replace them can have negative effects on team morale and productivity. "By driving those employees away, they've actually enhanced and sped up the very thing they were trying to stop," Ployhart said. https://lnkd.in/gWntKQF2

1