Angelo Mazzolini’s Post

Check out my Generative AI digital certificate issued by Data Masters! #digitalbadge #digitalcertificate #digitalcredentials #sertifier #applicationsofartificialintelligence #languagemodel #generativeai #text2image #agenticapplications

Check out my Generative AI digital certificate issued by Data Masters! #digitalbadge #digitalcertificate #digitalcredentials #sertifier #applicationsofartificialintelligence #languagemodel #generativeai #text2image #agenticapplications

When using GenAI for business purposes, don't fall victim to the hype that information provided by LLMs is equivalent to knowledge, because AI the label contains the word "intelligence". The root of true knowledge has always been and always will be context. This means LLMs must be helped. This requires the right data strategy and architecture. #AI #GenAI #LLM #architecture #data #strategy #knowledge #GPT-5 #ChatGPT #semantics The below video talks to the challenges.

Genie uses a compound AI system to interpret business questions and generate answers. Instead of using a single large language model, compound AI systems process tasks in AI applications by combining multiple interacting components. Compound AI systems are an increasingly common design pattern for AI applications because of their performance and flexibility. For more information, see The Shift from Models to Compound AI Systems. Reach us out for an Business Envisioning. #databricksgenie #anaxholdings #Anaxadvisory

Data Scientists & AI Developers should be able spend quality time in focusing on Idea/outcomes/GTM etc. RudraDB helps them with auto-intelligence feature.

Scale data engineering without scaling complexity DEaaS (Data Engineering as a Service) isn't about replacing engineers with AI – it's about giving them the support and smart defaults they need to deliver faster. Here's how it actually works 👇 #DataEngineering #DEaaS #AI #DataEngineering #Productivity #DataProductivity #TechLeadership

How to leverage Generative AI within Oracle APEX - It covers the integration of AI assistants and the seamless connection with your business data. The tutorial explores using services like OpenAI and Cohere to enable features such as Retrieval-Augmented Generation (RAG) for more accurate, context-aware AI responses. #OracleAPEX #GenerativeAI #AI #ArtificialIntelligence #Oracle #BusinessIntelligence #DataIntegration #LowCode https://lnkd.in/eg_eKmUq

SAS launches Academy for Data & AI Excellence in India to help bridge the country’s growing AI skills gap https://lnkd.in/en5GMzGA #referindia #referindianews #timesofindia #business #businessnews #NewsUpdate #TrendingNow #BreakingNews

I am pleased to share my recent project on predicting token usage in AI agent workflows—a key challenge in enterprise AI systems where costs scale directly with the number of tokens consumed. The idea for this project was inspired by an article from IBM on AI Agent Observability (https://lnkd.in/d3u9VA94), which highlighted how token usage has become a critical metric for monitoring and cost optimization in LLM-powered agents. Building on this concept, I generated a synthetic dataset to simulate enterprise observability data and designed a machine learning pipeline to forecast token consumption. The project involved detailed exploratory data analysis (EDA), feature engineering, and benchmarking of multiple regression models including Linear Regression, Random Forest, XGBoost, LightGBM, and CatBoost. The Gradient Boosting model emerged as the best performer (R² ≈ 0.84), accurately predicting token usage across diverse tasks. Key insights included the role of query length, number of tool calls, inference latency, and task type as the strongest drivers of token consumption. These findings show that token usage is systematic and predictable, enabling enterprises to forecast costs, optimize workflows, and scale AI systems more efficiently. The full project, including the report, notebooks, and code, is available on GitHub: https://lnkd.in/dzmDzdb3 Acknowledgement Gregg Lindemulder Annie Badman IBM Arvind Krishna Ruchir Puri Ryan Mandelbaum Jay Gambetta Jerry M. Chow Nikhil Gaddam Sangram Sinha Nickle LaMoreaux Surya Deep Singh Sandip Tripathy Sakina Mirza Sahil Desai Preethi Puthran Anjaneya Prasad Nidubrolu Vivek Kumar Tiwari Aryan Lala Malathi V Supriya Gawas Patade Mentor:- Praful Vinayak Bhoyar #AI #MachineLearning #DataScience #ArtificialIntelligence #LLM #AIAgents #PredictiveAnalytics #EnterpriseAI #CostOptimization #Innovation #Observability

https://lnkd.in/gpPr5f_e Reality check for #Enterprise AI engineering: Data pipeline that can’t deliver clean, contextualized, real-time inputs under governance constraints, your models will fail spectacularly. #DataScience #DataOps