Available for projects contract: active

Joan Antoni Morillo

Senior Data Engineer — Freelance

Reliable data platforms, and the AI tools on top of them

8+ years building pipelines, warehouses and cloud architectures on GCP and AWS. Lately I also build AI-powered tools for monitoring, analytics and ingestion.

Book a 30-min intro call Start a project See the work
  • 8+ years
  • 50+ pipelines
  • 4 AI products
datacontract.yaml 
dataContractSpecification: 1.1.0id: urn:datacontract:datawithjoaninfo:  title: Senior data engineer, freelance  owner: Joan Antoni Morillo  status: accepting_projectsterms:  usage: pipelines, warehouses, AI data tools  billing: fixed price | monthly retainerservicelevels:  intro_call: { duration: 30 min, cost: 0 }models:  engagement:    fields:      code_ownership: { value: client }      handover: { value: documented }

contract valid

About

Senior data engineering, without the overhead

One experienced engineer, from the first conversation to production.

I'm a Senior Data Engineer with over 8 years of experience designing, building, and running data systems for startups and enterprises. You work directly with me: no account managers, no handoffs.

My core is the full data lifecycle: ingestion, transformation, warehousing, and analytics, on GCP and AWS with dbt and Airflow. On top of that I build AI into the stack: natural-language analytics, automated document ingestion, agent-built pipelines, and incident diagnosis. Four of those products are in the Projects section below.

I'm taking on new freelance work. Tell me the problem, and I'll tell you plainly whether I'm the right person to solve it.

8+ Years of experience
50+ Pipelines built
4 AI data products shipped

Fit

Who I work with

Knowing who I am not right for saves us both time.

A good fit if you are

  • Growing teams on BigQuery, dbt, GCP or AWS whose data has outgrown its first setup
  • Companies with no data engineer who need a platform built, documented and handed over
  • Teams that want AI in their data tools, such as natural-language analytics or automated ingestion

Probably not a fit if you are

  • Looking for the lowest possible day rate
  • A full-time, in-office hire rather than a project or retainer
  • Web or app development with no data component

Ways to work together · engagement_offers

How I can help

Start small and de-risked, or bring me in for the long run.

Data Platform Audit

A clear picture of what you have and what to fix first.

fixed

What you get

  • Review of your pipelines, warehouse, costs and data quality
  • Quick wins you can ship immediately
  • A prioritised roadmap with effort estimates
Timeline
Typically 1 week
Pricing
Fixed price, agreed after a free scoping call
Talk about this

AI Pilot

One high-value AI use case, working on your data.

fixed

What you get

  • Natural-language analytics, document extraction or incident diagnosis
  • Guardrails, cost controls and an evaluation of the results
  • A working prototype and a go or no-go recommendation
Timeline
Typically 4 weeks
Pricing
Fixed price, agreed after a free scoping call
Talk about this

Ongoing Data Engineering

Senior capacity for the work that keeps piling up.

retainer

What you get

  • Build and run pipelines, dbt models and cloud infrastructure
  • Weekly demos and a shared backlog
  • Documentation and handover built in
Timeline
Monthly, no long lock-in
Pricing
Monthly retainer, scoped to your needs
Talk about this

What I work on

  • Pipeline Development
  • Cloud Architecture
  • dbt & Analytics Engineering
  • Data Warehousing
  • Data Quality & Observability
  • AI Integration
  • Data Governance
  • Data Strategy

Process

How an engagement works

Simple, predictable and easy to say yes to.

  1. 01

    Intro call

    A free 30-minute conversation about the problem. I tell you plainly whether I can help.

  2. 02

    Scope and plan

    A written scope with deliverables, timeline and a fixed price, so there are no surprises.

  3. 03

    Build

    Weekly demos, working in your repositories and cloud accounts, with you in the loop.

  4. 04

    Handover

    Documentation, runbooks and knowledge transfer, with optional support after launch.

stack · requirements.lock

Tools I use

Production-proven, chosen for the job.

Python
Python
SQL
SQL
Google Cloud
Google Cloud
AWS
AWS
dbt
dbt
Spark
Spark
Kafka
Kafka
Terraform
Terraform
Docker
Docker
BigQuery
BigQuery
LLMs
LLMs
RAG
RAG
MCP
MCP
···
& more

Selected work

Products I've built

Four data and AI products I designed and built end to end. Open a case study for the screens and how each works.

DataPulse case_study/datapulse-monitor

AI-Powered Data Platform Monitor

Built a full-stack monitoring hub that brings BigQuery, Kubernetes, dbt Cloud, GitHub, Datastream, Jira and more into one console. Cross-service incident correlation, on-demand AI root-cause analysis with Claude, schema-change detection with column-level lineage and blast radius, and a query advisor that flags cost anti-patterns with estimated weekly savings.

9 Platform integrations in one console
Claude AIBigQuerydbtKubernetesGitHubDatastreamJiraNext.js
Read case study
AI-Powered Data Platform Monitor screenshot
BQChatbot case_study/bqchatbot

Natural Language BigQuery Interface

Built a conversational interface to BigQuery where users ask questions in plain English and get back interactive charts and business insights. Claude generates optimized SQL via tool use, validates and dry-runs it for cost control, executes it against BigQuery, and auto-selects the best visualization from 11 chart types. A second AI pass writes headline insights, risks, quick wins, and anomalies from the results, with pinned dashboards, saved views, and scheduled alerts on top.

11 Chart types auto-selected by AI
BigQueryClaude AIReactNode.jsSSOPlotly.js
Read case study
Natural Language BigQuery Interface screenshot
Invoice Reader case_study/ai-invoice-reader

AI Carrier Invoice Pipeline

Built a pipeline that ingests carrier invoices in any format (CSV, Excel, PDF) from Drive, FTPS, SFTP, GCS, or REST APIs, extracts shipment records with Claude, and cross-checks them against the raw file (row counts, cost totals, tracking-code diffs) before loading into BigQuery. A React admin UI gives live per-carrier status, a failed/rejected file queue with retry, and a review queue for records the validator flagged.

5 Source types, one pipeline
Claude AIPythonBigQueryReactDockerGCP
Read case study
AI Carrier Invoice Pipeline screenshot
Datapipe case_study/datapipe

Agent-Driven ETL Pipeline Builder

Built a self-hosted ETL platform where AI agents do the integration work. Pick an app from a catalog of 31 pre-researched connectors (or name any API), choose endpoints and fields, and a codegen agent writes, test-runs and validates the pipeline into BigQuery, Postgres, Snowflake or DuckDB. Postgres CDC with resumable snapshots, drift detection, and an ops agent that diagnoses failures and proposes the fix, all in an operations console.

31 Pre-researched connectors
Claude AIPythonFastAPIReactBigQueryPostgres CDCSnowflakeDuckDB
Read case study
Agent-Driven ETL Pipeline Builder screenshot

Questions · faq.md

Before you reach out

The things people usually ask first.

How do we get started?
Book a 30-minute intro call. We talk through the problem, and I tell you plainly whether I can help and what a sensible first step looks like. No pitch, no obligation.
How is pricing set?
Audits and pilots are fixed price, quoted after a free scoping call so you know the cost before work starts. Ongoing work is a monthly retainer. I do not bill surprise hours.
Do you work with our existing stack?
Usually, yes. My day-to-day stack is GCP and AWS with BigQuery, Snowflake, dbt, Airflow, Python and SQL. If you use something else, tell me and I will say honestly whether I am the right fit.
We do not have a data team. Is that a problem?
No, it is a common starting point. I can design and build the platform, document it, and hand it over so a future hire or your existing engineers can run it.
Who owns the code and the work?
You do. Everything is built in your repositories and your cloud accounts, and it is documented so you are never dependent on me.
Can you sign an NDA?
Yes. Happy to sign yours before we discuss anything sensitive.
Why do the case studies use sample data?
The products are real and the screens match how they work, but the numbers and names are invented so that no company or customer information is ever shown.

Let's talk · POST /message

Tell me about your data problem

Describe what you're trying to fix, and I'll reply with honest next steps.

Prefer to talk?

A 30-minute intro call is the fastest way to find out if we're a fit: no pitch, no obligation.

Book a 30-min intro call
hello@joanantonimorillo.com
LinkedIn
GitHub