Statistics, Software, and AI Consulting

Make better decisions with AI built on statistical rigor.

Lucky Tech helps organizations turn complex data into reliable software, decision systems, and AI products—with the analytical depth to make them trustworthy.

Start a Conversation Meet the Founder

20+ years in statistics & technology Founder-led delivery End-to-end from decision to production

Services

For problems where a generic AI solution is not enough.

Work directly with a senior specialist who can frame the decision, validate the data, and build the system that puts the result to work.

Decision Intelligence

Statistical models, forecasting, experimentation, and data products for decisions that need to stand up to scrutiny.

Applied AI Systems

Practical generative AI, retrieval, automation, and ML systems designed around your domain, data, and risk profile.

Software and Data Platforms

Modern APIs, data pipelines, cloud architecture, and modernization that make complex systems easier to operate and evolve.

Technical Advisory and Enablement

Independent technical direction, team enablement, and practical training in statistics, AI, Python, and data systems.

Founder-led consulting

Depth before hype.

Lucky Tech is led by Démerson André Polli, a PhD statistician and software engineer. Every engagement receives direct senior attention—from problem framing through delivery.

Démerson André Polli, founder of Lucky Tech

A rare combination of quantitative and technical experience.

Démerson brings more than two decades across statistics, research, education, public-sector leadership, and software development. His work connects rigorous analysis with systems that people can use.

PhD and MSc in Statistics, with research in Bayesian methods and complex data analysis
16 years as a university professor and researcher at the University of Brasília
Former data and IT leader at IPEA and adviser in Brazil’s Ministry of Economy
First place ($10,000), Microsoft Innovation Challenge, November 2025 for AgroHelpdesk, a multi-agent service desk for agribusiness
Third place ($2,500), Microsoft Innovation Challenge, December 2024 for AI Search for Agricultural Planning and Control, an AI assistant for Brazilian agriculture
Second place, 2019 National Treasury Prize for fiscal-revenue forecasting research

Approach

From a difficult question to a system people can trust.

The work starts with the decision—not the technology—and combines statistical rigor, domain context, and engineering discipline through delivery.

Discover

Clarify the decision, stakeholders, evidence, constraints, and what a useful outcome looks like.

Architect

Define the data, model, software, security, and operating choices appropriate to the problem.

Build

Build production-ready pipelines, APIs, decision tools, and AI workflows with validation built in.

Scale

Monitor, document, and transfer the capability so your team can operate and improve it.

Expertise

Modern tools, grounded in quantitative thinking.

Python / Pandas scikit-learn / XGBoost TensorFlow / PyTorch BigQuery / SQL Vertex AI / SageMaker Power BI / Looker Airflow / dbt REST APIs / MLOps Bayesian inference Forecasting / experimentation Generative AI / RAG

Contact

Bring the hard problem.

Tell us about the decision, data, or system you need to improve. We will help you find a practical path forward.

Sao Paulo, SP, Brazil Global services available

Or connect on LinkedIn →