Software Engineer, Data Platform The data backbone behind AI dealmaking. Pipelines, warehouse, tooling. Engineering 4–6+ yrs Python · dbt · Airflow Postgres · AWS View role → →

Loop3studio; Buenos Aires; Hybrid; ARS 151,752,000 - 227,628,000; Full time

Job summary

Loop3studio is seeking a Software Engineer, Data Platform to own ingestion and data warehouse pipelines in a remote Argentina role. You’ll build end-to-end data flows, write dbt models, and maintain robust Python/SQL-based services that feed analysts and partners.

Expect collaboration with a small team and autonomous execution in a remote-first environment. You’ll work with technologies like Postgres, Airflow, Dagster, AWS/GCP/Azure, Docker, and modern data tooling, while shaping scalable

Qualifications

• Experience in Python and SQL for data engineering.
• Experience with Postgres and data warehouse pipelines.
• Familiarity with dbt, Airflow or Dagster, and cloud infra (AWS/Azure/GCP).
• Experience building data ingestion, warehousing, and APIs.

Responsibilities

• Build ingestion pipelines and end-to-end data warehouse flows.
• Develop dbt models and maintain data quality.
• Work with cloud infrastructure and containerized services (Docker).
• Collaborate with analysts and stakeholders to deliver reliable data products.

Skills

Python

SQL

Postgres

dbt

Airflow

Dagster

Docker

Cloud platforms

Kafka

DuckDB

LLM tooling

Tools

dbt

Airflow

Dagster

Docker

Kafka

DuckDB

LLM tooling

Job description

Engineering Remote · Argentina Paid in USD Full-time

Software Engineer, Data Platform

Top-50 VC firms decide where billions go. The signal under those decisions runs on pipelines someone has to build well. That's this job.

The ingestion + warehouse layer, end to end

Experience

4–6+ years backend or data engineering

Compensation

USD, discussed on call 1

Process

Python SQL / Postgres dbt Airflow / Dagster Cloud architecture & infrastructure (AWS, Azure or GCP) Docker Graph databases Semantic search Claude Code / Codex Nice: DuckDB, Kafka, LLM tooling

Fit check

Tick what's true. We'll tell you straight. Nothing is sent anywhere. This is for you, not us. Six statements, honest answer only.

Ingestion jobs, dbt models, internal APIs and the tooling that lets four other engineers move without asking you first.

PitchBook, Harmonic, Affinity, ten more. Reconciling entities that disagree about their own names and funding rounds.

With users

Calls with the people using what you built: analysts, ops leads, partners. You hear the complaint first-hand, not through a ticket.

Review

Reading other people's code and having yours read. No standup theatre.

T h r e e t h i n g s y o u ' d s h i p i n y o u r f i r s t y e a r

01 A signal pipeline a fund runs on daily

Dozens of sources into one warehouse, deduped and current. Sourcing team opens it every morning.

02 Entity resolution that stops the arguing

One company, five spellings, three funding numbers. You decide which one is true, and prove it.

03 Tooling that makes the next build a week, not a month

Shared contracts, tests, local dev that works. The multiplier on everyone else's output.

• Ambiguity is where you start, not where you stop.
• You care that the number is right, even when nobody would notice.
• You want the client relationship, not protection from it.

Be honest, probably not for you if

• You want a spec handed to you and a quarter to deliver it.
• Your favourite part of the job is the architecture diagram.
• Talking to the people who use your work sounds like a distraction.
• You need a large team's process around you to feel safe shipping.

W h a t y o u g e t

Same for every role, every level.

Remote-first

Work from anywhere in Argentina. If you live near Buenos Aires or Córdoba, you're in the office two days a month.

Performance bonus

On top of your salary, tied to how you perform.

Salary in USD

Paid in dollars, every month, reviewed twice a year.

Objective-based, with real adaptability. A few hours of overlap with US clients is the only fixed part.

14 days PTO + holidays

Taken, not accrued and forgotten. Local holidays on top, and the week between Christmas and New Year's off.

Working equipment

A MacBook Pro to work on, provided by us.

Tech budget

The tools that make you faster: Cursor Pro, Claude Pro, ChatGPT, whatever you'll actually use.

Courses, training and conferences that matter to the work. Ask, don't petition.

Written levels and what each one expects. Lead and staff tracks exist and are open.

High trust

Autonomy in exchange for excellence. Nobody checks whether you were online at nine.

What you've built, what you want next. We name our clients here.

Real problem, together

A messy dataset from actual work. No leetcode, no take-home over your weekend.

Founders + the team

Mutual diligence. Ask us the uncomfortable questions about money, clients and churn.

Then

Offer or a real reason, within a week of the last call.

Who you’d be working with

We keep you updated at every stage of the process.

Julian Iervasi Step 2 · the data interview

Nico Pisani Step 3 · the final call