Murali Krishnan

"Graphs make provenance first-class. Instead of burying "where did this come from?" inside a dense vector, a graph records sources, activities, and agents as nodes and edges you can query..." https://lnkd.in/efAdV9dx

26

"Principles of Building AI Agents" by Sam Bhagwat and found it incredibly insightful. The book provides a practical framework for understanding how to design, build, and scale AI agents covering everything from LLMs, prompting, RAG, MCP, multi-agent systems, and deployment strategies to the latest advancements in agent orchestration. As someone working on Agentic AI architectures, I found the sections on tool integration, agent memory, and dynamic agent workflows especially valuable for shaping enterprise grade solutions. Thank you Sam Bhagwat #AI #AgenticAI #AIArchitecture #MachineLearning #GenerativeAI #EnterpriseAI #RAG #MCP #Australia

21

2 Comments

2026 PIM Prediction: Bluemeteor's Agentic AI Revolution! Bluemeteor AI Labs leads PIM into agentic powerhouses with embedded AI Agents in PIM like Onboarding, Content Enrichment, Classification, unstructured product data extraction and Auto Match—achieving more than humanly possible! Shift from static repos to dynamic e-commerce drivers, evolving into "Service-as-Software" via autonomous workflows. AI Agents Evolve PIM -Automate onboarding, enrichment, validation; generate descriptions, images, videos, SEO optimization—80% less manual work. ​-Self-healing pipelines with QA agents, Multi-LLM, RAG, Vector DB for supplier ingestion, smart mapping, anomaly detection and more. Reshaping Markets & Service-as-Software -Goal-driven agents handle schema mapping, cross-channel syndication—minimal human intervention. -Feedback-learning cuts costs, accelerates product launches at scale from SKUs to millions. https://lnkd.in/gG3XDMNf Ready to scale? 💥 #AIAgents #PIM #Bluemeteor #AgenticPIM #EcommerceAI

18

1 Comment

Over the last few years, I have mentored founders across fourteen universities, twelve in Southern California and two in India, including my alma mater. The contexts were different. The ecosystems were different. The constraints were different. Yet the underlying need was the same. Founders everywhere are searching for clarity. Clarity on which problem is truly worth solving. Clarity on who the real customer is. Clarity on what to ignore, even when the opportunity looks tempting. In Southern California, I often meet founders who think global from the first day. They are ambitious and well connected, but many struggle to narrow their focus. In India, I meet founders who are remarkably resourceful and fast-moving, but who sometimes hesitate to expand their ambition beyond immediate markets. Neither approach is wrong. Each becomes stronger when balanced by the other. The founders who endure learn to pair vision with discipline. They resist the urge to chase every option and commit deeply to solving one meaningful problem well. Mentorship, in my experience, is not about prescribing direction. It is about helping founders develop their own internal compass, one that holds steady regardless of geography or environment. Across continents, the truth is consistent. Founders do not lack advice. They lack clarity.

90

8 Comments

#WinningwithAI in GTM isn’t about who builds it first -- It’s about who #DeploysitRight !! Before 2025 ends, 80%++ of customer interactions will involve AI. But here’s the irony: Most teams aren’t ready for it. Most AI go‑to‑market #GTM plans stumble not on tech, but on execution. 40%+ of AI projects stall because of data readiness issues. 65%+ of marketers say personalization is vital, AI can help. Imagine rolling out an AI tool. Teams get excited. Then… nothing moves :( :( The root? Dirty data. Disconnected systems. Misaligned teams. GTM AI success starts with clean data and clear goals. Without human oversight, AI can miss patterns or worse, amplify bias. Second mover strategy can work, learn from early missteps. Key lessons: #1. Start with a clear business problem, not just tech #2. Invest in data pipelines, monitoring and governance #3. Blend AI insights with human judgment #AILeadership #AI #GenAI #EnterpriseAI #AIGTM

23

ACTORS + AI: Director/Actor/Producer/Writer AI Chef 🧑‍🍳 Hans Obma visits the Cozy AI Kitchen 🎂 https://lnkd.in/gjiAvEfP to share his insights on how AI has transformed his creative processes. You may have seen Chef Hans in such 📺 hits as Better Call Saul, Wandavision, or Narcos — and yes, he's probably the most handsome AI 🧑‍🍳 chef we've had in the kitchen to date :-). Hans needed to learn a new language ... how did he do it? He got multimodal AI models to help him speak Welsh. Hans needed a way to scale how he communicated with his stakeholders ... how did he do it? He created an authentic-to-him way of scaling by creating an instant entourage of supporting agents. HT 🧑‍🍳 Ross Heise 🧑‍🍳 Matt Scholz and the power of the Wisconsin network for bringing such a storied talent to the Cozy AI Kitchen! --- This 🍰 episode: https://lnkd.in/gjiAvEfP All 50 episodes: https://lnkd.in/g6upvbGX

89

11 Comments

As AI evolves, responsible deployment is crucial. Building AI systems requires a focus on transparency, accountability, and fairness, especially when scaling across industries. - Transparency ensures AI is understandable and auditable. - Accountability requires clear ownership and human oversight throughout the system's lifecycle. - Fairness involves regular bias audits and testing to ensure equitable decisions. These principles are key to scaling AI responsibly, creating both business value and societal good. How is your organization approaching AI governance? What steps have you taken to ensure transparency and fairness?

629

5 Comments

A recent technical report from Beijing-based researchers at the Chinese Academy of Sciences, titled "SpikingBrain," is generating significant discussion. The paper introduces a family of brain-inspired large models, including SpikingBrain-7B, which reportedly achieves remarkable efficiency gains. Key performance claims from the report include: Inference Speed: The 7B model is said to attain an over 100-fold speedup in Time to First Token (TTFT) when processing long-context sequences of 4 million tokens. Power Efficiency: The model's design, which leverages an event-driven spiking mechanism, achieves 69.15% sparsity. This high level of sparsity is the basis for its claims of significantly reduced power consumption compared to traditional Transformer architectures. These results, if validated, represent a substantial advancement in efficient large-scale model design. As the paper is currently available as a preprint, the forthcoming peer review and independent replication of these findings are highly anticipated by the research community. Supporting Links: Primary Research Paper (arXiv): For a complete technical overview, you can access the preprint directly: https://lnkd.in/erfVy7Hh GitHub Repository: The project code and models have also been made available: https://lnkd.in/ejxD-5xJ

9

The simultaneous emergence of Claude Cowork and Clawdbot (now rebranded as Moltbot) is a clear signal of where consumer AI is headed this year: genuine end-user productivity through semi-autonomous agents. The days of purely prompt-based, one-shot results are giving way to systems that can handle real, ongoing work based on high-level requests. These tools are starting simple, but their trajectory is unmistakable: they will likely evolve quickly into capable digital assistants running 24/7. If recent AI trends hold, most of us will soon dedicate at least one (if not several) always-on machines or cloud instances to handling our routine personal and professional chores. This shift could free up significant time and mental energy, allowing us to focus on higher-leverage activities with far greater returns. Security, however, remains the single biggest hurdle. Granting AI agents access to local files, email, calendars, or messaging apps introduces real risks such as prompt injection, data leakage, unintended deletions, or worse. While tools like Cowork use sandboxes and careful permissions, and self-hosted options like Moltbot keep data local, vulnerabilities persist and demand vigilance. For now, the wisest path is cautious experimentation: start small, isolate sensitive data, monitor closely, and contribute feedback to help shape safer, more reliable agents. These early steps are opening exciting new chapters for everyday users. AI, it appears, won't just answer questions, but will truly work alongside us.

36

3 Comments