Karanveer A.

This blog post explores the shifting search landscape and how Vertex AI is best poised to meet developers in their journey of building powerful semantic, hybrid, and analytical search applications. Learn how Vertex AI Search out-of-the-box to DIY RAG APIs, empower developers to bring Google-quality search to ground applications in their data.

4

Processes with friction are sometimes a good thing. No, really. For example, my wife and I have avoided getting Amazon Prime because NOT having Amazon Prime introduces just enough friction to curb alluring consumerism. (That was until last week when we finally gave into Amazon Prime after 9 years of marriage). Don't be afraid to design friction into a process where appropriate and depending on the behaviors you need to incentivize. 2 examples: - Code Reviews are friction. This helps protect the code base and the customers. - Principle of Least Privileged Access. This is friction. It helps protect all sorts of data and systems. What other examples can you see friction being healthy in a process? --👇-- ♻️ Repost if you found this useful. ➡️ Follow me for more on strategic execution and leadership. 📨 Join my community newsletter, The TPM Craft (link in profile).

6

1 Comment

Really great lesson here in Slack vs Teams. Also I think this applies to Teams vs Zoom. Which would explain Zoom incredibly low stock growth over five years despite TONS of additional capabilities. Which best describes your product?

19

9 Comments

One of my krantikari tech leaders and ex-colleague Bhushan Deshpande discussed the possibility of using JHipster for accelerating the development of a product. JHipster is an incredible open-source platform that allows developers to generate full-stack web applications quickly, combining Java (Spring Boot) on the backend with Angular/React/Vue.js on the frontend. It’s a tool that has made life easier for many developers, but like any technology, it has its pros and cons. ✨ Why JHipster? - Seamless Full-Stack Integration: Quickly generates a full-stack app with best practices in mind, from frontend to backend. - Microservices Ready: Easily create and manage microservices architectures, supporting scalability and distributed systems. - DevOps Friendly: CI/CD pipelines, Docker, and cloud deployment are simplified, speeding up delivery. - Customizable & Extensible: Templates save time but remain fully customizable, allowing you to integrate your preferred tech stack. - Strong Community Support: With an active open-source community, JHipster is constantly improving, providing a wealth of plugins, tools, and integrations. ⚠️ Challenges to Consider: - Steep Learning Curve: Due to its full-stack nature and multiple options (Spring Boot, Angular/React/Vue), JHipster can be overwhelming for beginners. - Over-Generation of Code: Sometimes the generated code can be excessive or unnecessary, which might lead to complex maintenance if not carefully managed. - Heavy Resource Consumption: JHipster-generated apps, especially microservices, can consume a lot of memory and CPU, requiring more powerful hardware for local development. - Not Always the Best Fit for Small Apps: For smaller or simpler projects, JHipster’s full-stack capabilities can feel like overkill, and lighter frameworks might be more suitable. Despite these challenges, JHipster remains a powerful tool for building modern, scalable applications efficiently. As with any framework, it’s important to weigh the pros and cons based on your project’s specific needs. ⚖️ Have you used JHipster? What has been your experience? #JHipster #Java #SpringBoot #Angular #React #VueJS #DevOps #Microservices #FullStackDevelopment #OpenSource #SoftwareEngineering #TechChallenges #WebDevelopment

11

Can you trust AI? 🤔 For most companies the answer is no - at least using current foundational models out of the box. ❌ Problems like sensitive data leakage and toxic content are major barriers to implementing production genAI solutions. The route towards trust lies in system architecture and the good news is the field is advancing rapidly, including increasingly easily accessible toolkits of solutions to increase trust in AI systems. 🔥 This is the hot area for new releases from major AI platform players (along with AI security), including AWS (https://lnkd.in/e45S-ehS), DataRobot (https://lnkd.in/eCWvppWR), and Databricks (https://lnkd.in/e6A7cTcH). 👨‍⚖️ Guardrail or judge models - leveraging a second AI model to police the content coming out of an original - are probably the most promising scalable solution. For many the best route is likely to be a new class of foundational models specifically designed for this task such as Llama Guard and Databrick’s new judge LLM. There is also the ability to fine tune a model (think a discriminator model from a GAN architecture) or use a general purpose foundational model. 👮‍♂️ Classic filtering techniques for sensitive data can also be applied to the result of an LLM, whether for PII, company IP, or inappropriate content. This can include regex, classic PII models (such as Amazon Comprehend), etc. 🚻 Human in the loop techniques can also be used, but are far more scalable when used to calibrate automated solutions such as those described above rather than manually verifying outputs in realtime. What’s still missing? 🔎 🎯 While techniques to improve accuracy and reduce hallucination such as RAG and AI agents have made great strides, there remains a big gap in conveying a model’s confidence in its output to end users. 💡One idea is to leverage the probability distributions used for top-p (nucleus sampling) at the token level instead at greater aggregations to convey confidence levels for entire model outputs or significant segments (e.g. each step in a chain of thought structure or AI agents framework). #ai, #artificialintelligence, #GenAI, #GenerativeAI, #llm, #datascience, #machinelearning, #mlops, #bigdata

4

1 Comment

Is cache invalidation one of the toughest and most complex problems a software engineer can face in his career? (rhetorical question) If you are interested in the problem, I suggest reading the following piece from the Netflix Technology Blog. #software #engineering #cache #invalidation https://lnkd.in/dSW6_-8M

4