Thanks for the recording and hosting me 🙂
Read MoreSABANCI University: Restless Change of Enterprise Application Market – Impact of AI
Author: Ugur
Report: Meta Harness – Why we need it & Players available
What is a Meta-Harness & Why Do We Need It?What it is: A meta-harness is an orchestration and control layer that sits above individual AI agents to make them interoperable components of a unified system. It focuses on optimizing the executable support code around agent workflows—such as instruction files, setup flows, and validation scripts—rather than just the prompt. Why we need it:Harness Failures: Many agent failures stem directly from the harness, including weak repository instructions,…
Read MoreSAP Roadmap For Agent As A Service Landscape
A Massive New Frontier for Technical Partners SAP officially launched its “Autonomous Enterprise” vision on May 12, 2026, creating a massive new frontier for technical partners. This strategic pivot is backed by a dedicated €100M partner investment fund designed to accelerate ecosystem development. Industry analysis indicates that the Clean Core strategy has unlocked a highly lucrative agent-addressable market valued between €3-7B. This represents the largest platform-native agent opportunity currently available in enterprise software. Partners who…
Read MoreDatabricks Genie Ontology
The provided sources detail Databricks Genie Ontology, an automated, self-evolving context layer introduced in June 2026 to enhance enterprise AI accuracy. Built upon Unity Catalog, this system generates a “living knowledge graph” by extracting business logic from tables, dashboards, and connected applications. A core innovation is OntoRank, a PageRank-inspired algorithm that weights data definitions based on factors like author authority and usage frequency. This architectural shift significantly improves AI performance, raising text-to-SQL accuracy from roughly 50% to 84.5% on the first attempt. Unlike traditional, labor-intensive ontologies, this…
Read MoreThe Autonomous Enterprise Architecture for SAP and NemoClaw
These reports detail a massive shift in the enterprise AI landscape during mid-2026, centered on the collaboration between NVIDIA, SAP, and OpenAI. NVIDIA has introduced NemoClaw and OpenShell as a critical security and sandboxing layer for OpenClaw, the world’s most popular open-source AI agent platform. Simultaneously, SAP has integrated this technology into its “Autonomous Enterprise” vision, embedding hundreds of specialized AI agents into its core business software while blocking unmanaged agent access. OpenAI has expanded its enterprise footprint by launching the GPT-5.5 model on multi-cloud…
Read MoreNews-Based Trading ML Experiments – Technical Setup
Extracting Alpha from the News Cycle This documentation details a machine learning research project conducted in 2026 to evaluate how global news sentiment and economic calendars predict currency and commodity price movements. Researchers utilized the GDELT Global Knowledge Graph and economic event data to engineer fifteen unique features, testing them across 72 experimental configurations using gradient-boosted models like CatBoost and XGBoost. The study found that USDJPY was the most responsive instrument to news signals, with the highest-performing models achieving a Sharpe ratio of +2.653. Results indicated that a 24-hour prediction…
Read MoreAnalysis of ML Signals
https://ugurzafercandan.github.io/UTradeDashboard Open Orders list is showing a real time portfolio outcome. Majority of the signals are performing worse than anticipated back testing.I have results for 1 months of future test trading. Only 4 is showing positive returns. Rest is having issues in maintaining positive returns. I will continue monitoring
Read MoreSAP Joule Studio 2.0 & SAP Business AI Platform The technical blueprint
SAP Joule Studio 2.0 & SAP Business AI Platform The technical blueprint
Read MoreSAP AI Test & Demo Environments for Technical Audiences
SAP AI Test & Demo Environments for Technical Audiences
Read MoreThe Future of AI Agents – 8 Key Predictions
1- EVERYBODY Will Have Personal Agents 2- Company Memory Becomes an Asset 3- Different LLM Form Factors 4- LLM → Commodity / Agent/Harness → The Differentiator 5- APPS → Tools/Skills, USERS → Agents 6- Security/Governance Embedded in Harness 7- Agent Discovery & Collaboration 8- Testing/Hardening/Benchmarking Becomes Mainstream
Read More