Belinda Zeng, PhD

What are the most underrated Databricks capabilities? Watch my conversation with Shiv Trisal (Global Industrials GTM @ Databricks) on the topic below. For the full, practical, no-fluff discussion on real-world data + AI: https://lnkd.in/gMPsFHbY

89

One consistent challenge we see in the model training lifecycle is the lack of high-quality human preference data. Automated signals can drive scale but preference learning, reward modeling, and real-world model alignment still depend on expert creative feedback and structured human evaluation. Shutterstock delivers human-in-the-loop systems that generate aesthetic preference, human preference signals, and qualitative evaluation across training, fine-tuning, and model evaluation. This approach improves quality scoring, strengthens regression testing, and supports continuous learning, helping teams catch issues early and deploy models that reflect real human expectations. It’s how we support long-term partnership and ongoing collaboration with AI teams building production-grade systems. Contact us to learn more: https://lnkd.in/gtqmpYzy #ModelAlignment #ProductionAI #EnterpriseAI  #Shutterstock #FuelGreatWork

12

1 Comment

In the consumer world, we want a single interface that can write a poem, plan a vacation, and debug code. In that open world, scale is the only strategy that works. But enterprise workflows don’t live in an open world. Most B2B problems like parsing invoices, routing tickets, classifying clauses operate in closed systems. They have well-defined inputs, explicit outputs, and hard failure modes. In my latest article, I argue that the future of enterprise AI isn't always about getting bigger, it's often about model specialization. The most effective architectures I’m seeing are cascading. They use SLMs for routine volume, and escalate to LLMs only when deep reasoning is required. Read the full breakdown below. https://lnkd.in/gkyJbFWb

57

6 Comments

I hear it from customers every day - enterprises  that started their journey with AI by “experimenting” have evolved to a more holistic agentic approach that is driving real business value. Google Cloud's new ROI of GenAI study proves that AI is not about dabbling—it's about strategically deploying autonomous AI agents at scale. The data is clear: 📊 74% of executives report achieving ROI within the first year. 🤖 39% of executives have already deployed more than 10 agents across their enterprise. 📈 And among those seeing productivity gains, 39% have seen productivity at least double. And the impact of AI agents extends far beyond simple automation. In marketing, organizations are seeing 32% quicker content editing and 46% faster content creation. In security operations, AI agents are providing 70% reduction in breach risk and 50% faster mean time to respond to threats. Thank you Megan Poinski for covering the results in depth in Forbes, highlighting how companies are leveraging AI. You can read more there or download the study at: https://lnkd.in/gUYPwYuV https://lnkd.in/gfY3hG6Y #AI #GenerativeAI #GoogleCloud

440

24 Comments

Dr. Ramla Jarrar, an MMM expert, shares an important perspective every marketer using MMM should read. The customer journey is complex, and while MMM is a powerful tool for understanding how spend impacts that journey, its limitations are often overlooked and nuances brushed over once results are packaged in polished decks. The tension I often see: marketers need one strong KPI to track large marketing investments, but the complexity of measurement means no single framework provides that holy grail metric. Measurement methodologies are sophisticated and technically robust, yet each has constraints. Triangulating across multiple frameworks—MMM, incrementality testing, attribution, brand tracking—provides better insights, though interpretation remains as much art as science. This requires partnership. Marketers need to understand what each methodology can and cannot deliver. Analytics partners, whether internal teams or external consultancies, have the responsibility to explain assumptions, acknowledge limitations, and help interpret results responsibly. When both sides engage with this complexity honestly, we make better decisions. A question for marketers: where do you find the biggest gaps between marketing performance measurement and the questions you need answered? #MarketingMixModeling #MMM #MarketingAnalytics #MarketingMeasurement #MarketingROI #DataDrivenMarketing #MarketingStrategy

1

AstroBee generates ontologies...but what is an ontology, and if they're important, why don't you already have one? I've written a short piece on Kimball star schemas, ontologies, and how each relate to Bill Inmon's original approach to data warehousing in the early 90s. Dig in to find an overview of a fascinating technological debate and the re-emergence of an approach to data modeling that's been sleeping for 20 years.

18

3 Comments

Last week I spoke at Snowflake's World Tour event. The data ecosystem + career paths have changed quite a bit in the last four months since I presented at Snowflake Summit. A few changes I've seen on our data team since then: Today’s AI tools shorten "time-to-impact". We've seen new hires use Claude Code, Cursor, Notion AI, etc. to onboard faster to codebases and business domains. If you can search, reason, and ship, you can contribute far faster than before. You cannot measure agentic product experiences without full-stack data talent. Ramp's proportion of LLM products, questions, and decisions will only go up. AI-as-a-feature requires eval datasets, offline/online metrics, and drift monitoring. This is product analytics with semantic rigor vs. siloed DS research or SQL-only data modeling. Increasingly, the skill we're asking our team to develop / hiring for is owning the full product loop (metrics → modeling → experiment → decision → monitor) vs. deep specialization in one tool, language and ecosystem (whether "applied science" or "analytics engineering"). Big thanks to Nicole and Bobby for having me!

114

4 Comments