Reconcile scattered sources and make every transformation inspectable.
Portfolio — Thomas Brochard
Design & developmentFrom raw data
to useful systems.
I connect data engineering, development and AI to make complex systems clearer and easier to use.
01 /Selected projects
Design, put into practice.
Concrete challenges.
Considered technical decisions.
Connect terrain, weather and versioned changes in a mission-planning design.
Further explorations
A collection of 10 case studiesWeather ELT Pipeline
Normalize heterogeneous weather feeds into traceable analytical observations with layered SQL and explicit quality rules.
CSV to MongoDB Migration
Validate flat files, map typed documents and track bounded batch imports with checkpointed progress.
Sales Warehouse Audit
Connect retail revenue reconciliation, warehouse integrity checks and SQL change-capture concepts.
Building Energy Prediction
Structural feature engineering, comparable regression workflows and a typed prediction contract for building-energy estimates.
Education Market Analysis
Turn sparse education indicators into comparable country profiles and transparent scoring scenarios.
Real Estate SQL Analytics
A normalized transaction model for geographic comparisons, temporal analysis and grain-aware SQL queries.
Cultural Events RAG
Connect natural-language questions to cultural events through metadata-aware retrieval, grounded answers and source references.
RAG Platform Architecture
A proposed cloud architecture for session-aware retrieval, catalogue refresh, observability and cost planning.
These are technical demonstrations, selected-source showcases and architecture studies. The case studies describe their scope and limitations. GitHub repositories are private and require authorized access.
02 /Method
Complexity deserves
a clear structure.
A project is understood through its decisions as much as its interface. Each of those decisions deserves to be explicit.
Define the right boundaries.
Clarify the need, available data and scope before choosing the tools.
Build reliable connections.
Define the models, contracts and checks that connect the components of a system.
Make decisions readable.
Document the reasoning, expose the limits and design understandable interfaces.
03 /About
The bigger picture.
An eye for detail.
I’m interested in what connects data to its uses: its structure, how it moves and the way we interact with it.
This portfolio brings together my work on data pipelines, analytics, retrieval systems and interfaces. Each study makes the technical reasoning and the project’s limits visible.
My GitHub profile04 / Areas of practice
From the interface
to the data flow.
My work connects what a person sees with the models, transformations and workflows that support it.
01Interfaces with clear state.
Structure, responsive layouts and interaction design. In complex interfaces, that also means preserving drafts, explaining unknown states and making concurrent edits visible.
In practice: Fly Drone Tool ↗
02Workflows with explicit steps.
Python scripts, orchestration and batch processes. Make dependencies inspectable, keep progress visible and place quality checks where they can explain a failure.
In practice: BottleNeck ↗
03Data with a traceable path.
Relational models, SQL transformations and analytical layers. Define the grain, preserve provenance and reconcile records before drawing a conclusion.
In practice: Weather ELT Pipeline ↗
The next conversation