Senior Analytics Engineer

I turn business questions into reliable numbers.

Senior Analytics Engineer based in Paris, France. Five years designing and owning analytics platforms end to end, from cloud architecture and dbt modeling to BI delivery and business adoption. What I deliver is measured in decisions enabled, not models deployed.

Book a call
Get in touch
Thierry Triet LAM, Senior Analytics Engineer
Paris, France

Nine companies across four countries, in consulting and product teams

Selected companies

Track record

Years of experience
5+
Companies across 4 countries
9
SQL generation time at Connecty AI
−80%
Annual revenue loss cut by 92% with a single project
€2M

Value proposition

What I bring to your team.

Four business domains owned end to end, backed by platform craft and AI leverage. Every card is a shipped result, not a promise.

Finance

Payment reconciliation

Multi entity, multi currency payments consolidated into one Finance mart, safeguarded by reconciliation and automated quality tests.

35 Stripe accounts · 1 Connect platform · 6 entities

Supply chain

Stock distribution

The Refused Store project: OMS, Salesforce and SAP crossed to expose in store order refusals and drive corrective action.

€2M annual revenue loss, cut by 92%

Marketing

Lead attribution

CRMs and marketing spend matched into one attribution model, with funnel models that follow every euro from acquisition to revenue.

Salesforce · Pipedrive · GA4 · Meta Ads, unified

Sales

Omnichannel reporting

International sales consolidated with forex conversion and order lifecycle control: one trusted view of group performance.

4 luxury brands · group wide KPIs

Platform

Rebuilds, from scratch or legacy

Three platforms rebuilt in three contexts: solo greenfield, a legacy monolith broken down, and a startup stack stood up from nothing.

3 platforms · 225 models solo in production

AI & LLM

AI accelerated engineering

AI agents run in parallel to analyze context, verify, develop and test automatically, with Claude Code at the core of the workflow.

Claude Code daily · parallel agents · LLM benchmarking

About

Owned end to end, measured in decisions.

Five years designing and owning analytics platforms end to end, from cloud architecture to the dashboards leadership actually opens: nearly 60 dbt data projects for 9 companies across France, the UK, the US and Germany.

What you get is a platform that is tested, documented and self service, so your team spends less time fixing data and more time deciding with it. Building or fixing a data platform? I would like to hear about it.

Location
Paris, France
Experience
5+ years
First contact
Book a call
Languages (fluent)
French, English, Vietnamese

Footprint

9 clients · 4 client countries · 35 countries traveled

World map: 9 clients · 4 client countries · 35 countries traveled
Map © svg-maps · CC BY 4.0

Stack

Skills and tools

How I scope, deliver and land analytics platforms: the delivery practices first, then the stack that serves them.

Delivery & business impact

  • Business requirements scoping
  • Product Vision
  • Roadmap
  • KPIs & reporting
  • Variance analysis
  • Business adoption
  • Prioritization
  • Documentation
  • Jira
  • Confluence
  • Notion

Core stack

BigQuery

4 years · Medallion at scale · Data Mesh at Kering

dbt

~1,000 models across 60 projects · Medallion · 3NF

SQL

A 3,000 line monolith refactored · star schema · data marts

Tableau

LOD · the Refused Store dashboard · quality monitoring

Power BI

DAX & Power Query · migrated into dbt, versioned and tested

Snowflake

Full stack from zero · LLM semantic layer · resource monitors

AWS

Athena · Redshift · S3 · a dbt platform migrated to GCP

Git & AI

Claude Code daily · parallel AI agents · LLM benchmarking

From sources to decisions

One pipeline, from sources to decisions.

The reference architecture I deploy on every mission, on GCP or AWS: business systems land through managed connectors into a governed warehouse, are modeled in dbt, and ship as dashboards each team actually uses.

01Sources

SAP S/4HANA ERP · Finance
OMS Manhattan Order management
Okaneo PIM · Product
Salesforce CRM
Stripe Payments
GA4 Web analytics
Amazon S3 Files · object storage

Also connected: Pipedrive · Shopify · Typeform · Airtable · Meta Ads · Search Console

02Ingest

Fivetran Managed EL
Airbyte Open connectors

Incremental syncs · Python & PySpark for the long tail

03Transform

dbt Modeling & semantic layer
Raw
Staging
Intermediate
Marts
~1,000 models · 60 projects139 SCD2 snapshotsCI-tested on every PR
Governed warehouse & lakehouse
BigQuery Snowflake Databricks

04Serve

Tableau LOD · exec dashboards
Power BI DAX · Power Query
Data Studio Self service reporting

05Decide

Finance Payments reconciled to the cent
Marketing Every lead tied to revenue
Supply Chain Stock aligned with demand
Sales 4 brands · forex consolidated

Orchestration & CI/CD

  • Airflow
  • dbt Cloud
  • GitHub Actions
  • Terraform
  • Docker

Quality & observability

  • Freshness & volume tests
  • Reconciliation to source
  • Monitoring dashboards

Governance & FinOps

  • Documentation & lineage
  • IAM & access
  • Warehouse cost control

One warehouse and one BI tool per mission: this is the superset I adapt, proven at Colonies, Kering and Connecty AI.

Career

Professional experience

Six roles since 2021, consulting and permanent, each told through the change it made.

Since March 2026 · Solo

Senior Analytics Engineer Consultant

Colonies · Real estate · PropTech

Solo rebuild of the analytics platform of a European real estate operator across four countries: eight fragmented source systems turned into one tested BigQuery foundation covering Finance, Ops, Sales, Marketing and Asset Management.

225

production dbt models, owned solo, across five business domains

35

Stripe accounts under one Stripe Connect platform account (6 entities, 4 countries), reconciled in the Finance Data Mart

139

SCD2 snapshots tracking history, plus 16 automated quality tests

Approach

  • Payments consolidated: 35 Stripe accounts under one Connect platform account reconciled, with fund routing audited through the Express to Standard migration.
  • Acquisition cost and rental revenue tracked: Salesforce, Pipedrive and marketing spend matched, lease lifecycle and occupancy modeled as reusable dimensions.
  • Reporting industrialized: dbt Cloud pipelines orchestrated, 139 SCD2 snapshots for history, Data Studio rebuilt as self service.
  • dbt Cloud
  • BigQuery
  • Data Studio
  • Fivetran
  • Stripe
  • Salesforce
  • Pipedrive
  • SQL
  • GitHub

Dec 2025 to Mar 2026 · Solo

Founding AI Data Analyst

Connecty AI · Agentic AI startup · Remote, USA

First analytics hire at an agentic AI startup: stood up the internal data stack (Snowflake, dbt, Power BI) and validated the LLM natural language to SQL engine.

−80%

SQL generation time, cut by the semantic layer restructure

3 axes

LLM benchmark (accuracy, latency, cost) that guided the engine choice

Power BI (DAX) → dbt

measures extracted from TMDL files into versioned, testable models

Approach

  • Generation time reduced by 80%: semantic layer structure and marketing funnel templates optimized.
  • Acquisition-to-revenue chain unified: Facebook Ads, GA4 and Shopify centralized in dbt.
  • Cloud budget controlled: warehouse sizing, query profiles and resource monitors catching cost drift before billing.
  • Snowflake
  • dbt
  • Power BI
  • Semantic layer
  • GA4
  • Facebook Ads
  • Shopify
  • LLM prompt engineering

Feb 2023 to Dec 2025 · Team of 14 · Digital Business Intelligence (DBI)

Senior Analytics Engineer Consultant

Kering · Luxury retail

Inside Kering Technologies, I helped modernize the group analytics platform for global reporting across Sales, Omnichannel, Consumer Journey, Supply Chain and Marketing, and my own initiative, the Refused Store project, turned stock distribution insight into recovered revenue.

€2M

annual revenue loss (2022 global sales): cut by 92% with the Refused Store project

3,000

line SQL monolith refactored into modular, tested dbt models

4

brands consolidated with forex conversion and order lifecycle control for group wide KPIs

Approach

  • Revenue loss reduced by 92%: launched the Refused Store project on my own initiative, a Tableau dashboard crossing OMS, Salesforce and SAP on in store order refusals.
  • Platform migrated to BigQuery: the dbt project moved off AWS Athena and S3 under a Medallion architecture, applying the group's Data Mesh strategy.
  • Data quality industrialized: dbt checks for freshness, volume thresholds and parameter consistency, plus a Tableau dashboard letting business teams validate the metrics themselves.
  • BigQuery
  • dbt
  • AWS (Athena, Redshift, S3)
  • PySpark
  • Airflow
  • Airbyte
  • Tableau
  • Power BI
  • Terraform
  • Docker
  • Agile (Scrum)

Aug 2022 to Jan 2023 · Team of 3 in Paris + offshore ops in Mumbai

CRM Data Architecture Consultant

Accenture · Automotive CRM · Jaguar Land Rover

International CRM transformation program for Jaguar Land Rover: SAP S/4HANA and Salesforce Marketing Cloud integrated to serve sales, customer service and campaigns across five continents.

5

continents served by the CRM data integration

95%

incident resolution rate with SAP R&D through go live and hyper care

Approach

  • CRM data reliability at international scale: SQL extraction and segmentation pipelines automating de duplication, normalization and cleansing.
  • Stakeholder and incident management: 95% resolution rate with SAP R&D, stabilizing operations through go live and hyper care.
  • SAP S/4HANA
  • Salesforce Marketing Cloud
  • SQL
  • Tableau
  • UML
  • Data integration & ETL

Feb 2022 to Jun 2022 · Team of 6

Data Governance Engineer

Veepee · E commerce

Post acquisition integration: unified the data of acquired companies into one centralized cloud warehouse, with shared analytics layers and governance.

1

governed, centralized warehouse unifying every acquired company

End to End

quality and governance rules across the whole data lifecycle

Approach

  • Business-oriented dbt models: transformation models delivering business ready datasets to the teams.
  • Data quality and governance: rules established across the full data lifecycle, ensuring reliability.
  • dbt
  • BigQuery
  • SQL Server
  • GCP
  • Airflow
  • Python
  • Docker
  • GitLab

Apr 2021 to Dec 2021 · Team of 6

Data Architecture Engineer

Dassault Systèmes · Industrial software · R&D

R&D role on the 3DEXPERIENCE platform, owning the data architecture of scalable SaaS systems.

Data lake

data models, metadata frameworks and storage layers: its groundwork

WASM

components integrated via Emscripten, aligned with application constraints

Approach

  • Data lake foundations: data models, metadata frameworks and storage layers engineered for scalable SaaS systems.
  • Data and application integration: components integrated through WebAssembly and Emscripten, aligning the data system with application constraints.
  • SQL
  • XML
  • JSON
  • C++
  • WebAssembly
  • Emscripten
  • JavaScript
  • Linux

Credentials

Certifications

Continuous learning across the modern data and AI stack.

Hands on work

Selected projects

Real engagements, rebuilt on synthetic data and told as data stories. Each one starts with a business question and ends with a decision, with the artifacts on show. The experience section names the company. Projects let you explore the method.

Cover for the project: Which salespeople missed their target? Synthetic data

Which salespeople missed their target?

Why did some salespeople miss their quarterly target when the company beat plan overall?

  • dbt
  • DuckDB
  • SQL
  • Python
Read the data story
Cover for the project: Is the new build safe to ship? Synthetic data

Is the new build safe to ship?

Should build 11.0 go to a full rollout, and if not, what exactly broke?

  • Python
  • DuckDB
  • SQL
  • matplotlib
Read the data story

Path

Timeline and education

  1. Since March 2026 · Paris, France

    Senior Analytics Engineer Consultant (Data Platform & Reporting)

    Colonies logo Colonies

    Current consulting mission: sole owner of the whole analytics estate, from ingestion to the reporting layer, for a European real estate operator.

  2. Dec 2025 to Mar 2026 · California, USA (Remote)

    Founding AI Data Analyst

    Connecty AI logo Connecty AI

    Founding role held remotely from Paris for a California startup, hired to make analytics exist before the product shipped.

  3. Feb 2023 to Dec 2025 · Paris, France

    Senior Analytics Engineer Consultant (Data Platform)

    Kering logo Kering

    Longest mission to date: three years inside Kering Technologies, serving group reporting for the houses through the move to Data Mesh.

  4. Aug 2022 to Jan 2023 · Paris, France

    CRM Data Architecture Consultant

    Accenture logo Accenture

    Consulting delivery through a go live: keeping Jaguar Land Rover's CRM data operations stable alongside an offshore team.

  5. Feb 2022 to Jun 2022 · Paris, France

    Data Governance Engineer

    Veepee logo Veepee

    First role centered on governance, inside the central data team of the e commerce group during its acquisition wave.

  6. Apr 2021 to Dec 2021 · Vélizy-Villacoublay, France

    Data Architecture Engineer

    Dassault Systèmes logo Dassault Systèmes

    R&D on the 3DEXPERIENCE platform. Data models, metadata frameworks and storage layers for the data lake. WebAssembly integration.

  7. 2020 to 2021 · Poitiers, France

    Master of Software Engineering in Big Data & Embedded Systems

    ISAE-ENSMA logo ISAE-ENSMA

    Master focused on Big Data and embedded systems.

  8. 2015 to 2021 · France · Vietnam

    Double Aeronautical Engineering Degree in Data and Systems

    ISAE-ENSMA logo ISAE-ENSMA
    Ho Chi Minh City University of Technology logo Ho Chi Minh City University of Technology

    Eiffel Excellence, ERASMUS+ and SEED fully funded scholarships. Second prize, Vietnam Mathematics Contest 2017.

References

What the people I reported to say

Quoted verbatim from written recommendation letters. Each is identified by role and company; a quote written in the other language keeps its original wording, with the translation underneath.

Reference letters are available upon request and can be shared for specific position applications.

Colonies Tech Lead Colonies Direct manager
He was not simply keeping the lights on; he improved what he inherited. Any organisation looking for a data engineer who can own a modern stack and speak credibly to the business at the same time would be fortunate to have him.

1/4

Kering Technologies Lead, Data Analytics Kering Technologies Direct manager
His initiative has helped the commercial teams make data-driven decisions to optimize sales processes. His ability to combine technical rigor with business understanding makes him a valuable asset to any data team.

2/4

Accenture Song Delivery Market Lead, Senior Manager Accenture Song Direct manager
Always focused on the task at hand, quick and diligent and with a very pleasant personality. Any venture who will consider Triet as an employee will get an excellent individual who is hard working with profound data science knowledge that will strengthen the team's performance.

3/4

Dassault Systèmes R&D Software Engineering Director Dassault Systèmes Manager's manager (N+2)
Minh Triet LAM est très volontaire et organisé dans les missions qui lui sont confiées. Il va de l'avant et surmonte les difficultés avec persévérance.

Translation: Minh Triet LAM is highly committed and organised in the assignments entrusted to him. He moves things forward and overcomes difficulties with perseverance.

4/4

Contact

Turn your data into decisions you trust.

Freelance missions or a permanent role, in Paris or remote. The fastest way to reach me is email or LinkedIn.

Book a call Email me