Talk-to-Data
-
- Governance-first semantic layer + LLM chatbot for quotation
- analysis. Guardrails live in dbt transforms, not prompt
- engineering. Cortex masks sensitive data by design.
+
+ I’m Gabriel Pereira. I turn complex business questions into governed data foundations, thoughtful architecture, and solutions people can actually use.
Architecture · Analytics · Applied AI Good architecture connects technical rigor to the decisions a business needs to make. Reliable by design Governed & semantic Useful beyond the demo
+ I’m a Data & Analytics Solution Architect with nine years at Siemens, working where business context, data engineering, and applied AI meet.
+
+ My experience spans governed AI foundations, semantic layers, forecasting, and production-minded data products. I care as much about the decisions behind the system as I do about the system itself.
+ Case studies built to be inspected—not just admired. Architecture, governance, trade-offs, and reproducible implementation. A governance-first semantic layer and LLM assistant for quotation analysis. Guardrails live in dbt transformations—not in hopeful prompts—with sensitive data protected by design. A baseline-to-model delivery flow with rolling-origin validation. Stakeholders can audit the familiar Excel baseline before the gradient boosting model is even trained. A product knowledge assistant for HVAC specification questions, combining retrieval, Gemini embeddings, and measured performance in a production-minded MVP. Smaller tools, data pipelines, automation helpers, and exploratory prototypes. An evolving workspace where new ideas become reusable delivery patterns. A new home for practical notes, technical walkthroughs, and honest lessons from building with data and AI. Field notes on data architecture, analytics engineering, and applied AI—what works, what breaks, and what I learn along the way. Technical walkthroughs, practical experiments, and conversations about building useful things with data, analytics, and AI. Four focused offers for teams and professionals who need clarity, practical capability, or software built around the way they actually work. Purpose-built tools and data products designed around your operation—not the other way around. Technical clarity for teams navigating data architecture, analytics, AI adoption, or a complex delivery decision. A structured view of where data and AI can create value in your operation—and where to start first. Practical, on-demand sessions tailored to the context, tools, and goals of professionals and teams—building confidence and autonomy through real application. These are upcoming independent services. For current career goals and opportunities, connect with me on LinkedIn ↗. Whether it’s a professional opportunity, a technical idea, or a challenge you’re trying to understand—I’d be glad to hear from you.Making data useful.
+
Making AI practical.
+
+ Gabriel Pereira
+ From complexity to clarity.
+ Data foundations
+ Trusted products
+ Decisions & AI
+ Technology is the means.
+
Better decisions are the point.
+ “Make it trustworthy. Make it understandable. Then make it intelligent.”
+
+ Open work.
+
Real engineering decisions.Talk-to-Data
+ Demand Forecasting
+ HVAC RAG Assistant
+ Workspace Experiments
+ Sharing the process,
+
not just the outcome.Notes from inside
+
the data factory.Watch ideas become
+
working solutions.A new way to work
+
together is taking shape.Custom Software
+
+
+
+ Register your interest
+
+
+ Consulting
+
+
+
+ Register your interest
+
+
+ Opportunity Map
+
+
+
+ Register your interest
+
+
+ Data & AI Training
+
+
+
+ Discuss a training session
+
+
+ Good work starts with
+
a good conversation.