Jobs › Companies › Avra › Member of Technical Staff | Knowledge Platform

Über diese Member of Technical Staff | Knowledge Platform Stelle bei Avra

Avra · Remote · São Paulo

About Avra

Avra is a frontier AI lab building relational foundation models from the ground up (no third-party wrappers).

We model how businesses and individuals are connected and interact over time. Put simply: we bring connections into the math.

Leading enterprises in LatAm use Avra's models to power key decisions in credit, growth, and fraud. Our customers include the largest banks and marketplaces in Latin America, such as Mercado Libre, Telefonica, Banco Inter and Getnet.

Avra was founded in 2024 by Viviane Meister and Bruno Alano (ex-OpenAI). Our investors include NXTP, Clocktower Ventures, as well as the Sequoia and a16z scouts. We are accelerated by NVIDIA, AWS, and Google.

We are a team of builders. Join us.

About the role

Avra is a frontier AI lab building relational foundation models. You own how data gets into Avra and what shape it is in when it arrives.

That covers two paths. The first is acquisition: crawlers, scrapers and browser automation that collect data from sources with no API, and keep collecting when those sources change. The second is the data platform itself: the streams, pipelines and lakehouse that turn raw captures and large relational datasets into versioned tables that research, training and production predictions depend on.

This is a hands-on role for someone who has done both sides: built collection systems against difficult sources, and moved large volumes of data through production pipelines.

What you'll do

  • Build and run the acquisition layer: crawlers, scrapers and browser RPA that collect data from websites and portals, with scheduling, retries, session handling and change detection.

  • Keep collection reliable as sources change: detect layout and access changes early, track coverage and freshness per source, and repair quickly.

  • Run event-driven ingestion on Kafka and Pub/Sub, including ordering, replays, dead-letter queues, backpressure and idempotent consumers.

  • Build batch and incremental pipelines in Spark, dbt and SQL that turn raw captures into clean, typed, relational tables.

  • Own the lakehouse on Parquet and Iceberg: schema evolution, partitioning, compaction, and snapshots that make training sets reproducible.

  • Set data quality and lineage standards: contracts, tests, and traceability from a prediction back to the source record.

  • Operate everything on Kubernetes, packaged with Helm, provisioned with Terraform and shipped through CI/CD with reliable rollback.

  • Collect responsibly, in partnership with Security and Legal: rate limits, terms of use, LGPD and handling of personal data.

How we measure success

  • Freshness: time from a change at the source to that data being available in the lakehouse.

  • Coverage and success rate per source, and time to recover when a source breaks.

  • Time to onboard a new source, from first request to production collection.

  • Pipeline reliability: on-time runs, failed runs and consumer lag.

  • Data quality: issues caught by tests before consumers see them, against issues noticed downstream.

  • Cost per record collected and per terabyte processed.

What we're looking for

  • 5+ years of relevant engineering experience at tech and product-driven companies.

  • Required: a production background in at least one of web scraping, large-scale crawlers, or browser RPA. You have built and operated collection against real sources, not only consumed APIs.

  • Required: a data engineering background with large volumes of data, using Spark, dbt and SQL, on columnar and table formats such as Parquet and Iceberg.

  • Experience with messaging and streaming systems in production: Kafka and/or Pub/Sub.

  • Production code in Python and JavaScript, from pipelines and crawlers to browser automation.

  • Hands-on Kubernetes experience, including packaging with Helm.

  • Infrastructure as code with Terraform, and CI/CD pipelines with reliable rollback.

  • Willingness to operate what you build, including incident response when a source or pipeline breaks.

  • Daily use of coding agents, with a considered view on using them safely.

Nice to have

  • Collection from difficult sources: heavy JavaScript rendering, logged-in sessions, rate limits and proxy management, with tools like Playwright, Puppeteer or Scrapy.

  • Workflow orchestration with Airflow, Dagster or Temporal.

  • Entity resolution and relational or graph data modeling across many sources.

  • Datasets for ML: point-in-time correctness, reproducible training sets and evaluation splits.

  • GCP and AWS data services, including GKE, BigQuery, Dataproc and EKS.

  • Observability with OpenTelemetry.

  • Financial services or regulated environments, including LGPD and ISO 27001.

Bereit, sich bei Avra zu bewerben?
Bei Avra bewerben

Ähnliche Jobs

Avra
Member of Technical Staff | Deployment Platform
Avra
⚡ Früh bewerben São Paulo · standortgebunden
● Neu 👁 Gesehen ✓ Beworben vor 1 Wo.
Avra
Member of Technical Staff | Product Platform
Avra
⚡ Früh bewerben São Paulo · standortgebunden
● Neu 👁 Gesehen ✓ Beworben vor 1 Wo.
Avra
Member of Technical Staff | Inference Platform
Avra
⚡ Früh bewerben São Paulo · standortgebunden
● Neu 👁 Gesehen ✓ Beworben vor 1 Wo.
Avra
Deployment Strategist
Avra
⚡ Früh bewerben São Paulo Hybrid
● Neu 👁 Gesehen ✓ Beworben vor 2 Wo.
Avra
Open Application
Avra
⚡ Früh bewerben São Paulo · standortgebunden
● Neu 👁 Gesehen ✓ Beworben vor 3 Wo.
Avra
GTM Enterprise - Founding Team
Avra
⚡ Früh bewerben São Paulo Hybrid
● Neu 👁 Gesehen ✓ Beworben vor 3 Wo.
Avra
Member of Technical Staff, Research — Relational Foundation Models
Avra
⚡ Früh bewerben Remote · standortgebunden
● Neu 👁 Gesehen ✓ Beworben vor 3 Wo.
Avra
Head of Research
Avra
⚡ Früh bewerben Remote · standortgebunden
● Neu 👁 Gesehen ✓ Beworben vor 3 Wo.
Rws
Brazilian Portuguese - Religion & Philosophy Subject Matter Expert (SME)
Rws
⚡ Früh bewerben São Paulo · standortgebunden
● Neu 👁 Gesehen ✓ Beworben vor 2 Std.

Registrieren für Vorschläge, die auf die von Ihnen geöffneten Jobs und gespeicherten Suchen zugeschnitten sind.

Mehr Jobs bei Avra

Alle Jobs bei Avra ansehen →

Jetzt bewerben
🤖

Moment — langsam

JobsRadar wurde für echte Menschen gebaut, die eine schwere Zeit bei der Jobsuche haben — nicht für automatisierte Anfragen. Sie klicken viel zu schnell und sind jetzt vorübergehend blockiert.

Kommen Sie später wieder. Wenn Sie wirklich auf Jobsuche sind, stehen wir hinter Ihnen — verhalten Sie sich einfach wie ein Mensch.

Catch your next role the second it’s posted.

Create a free account and we’ll watch the boards for you — the instant a job matches your search, it lands in your inbox or Telegram. No digging, no refreshing.

Create free account

Free forever · takes 30 seconds · already have one?

Verschaffe dir einen Vorsprung bei der Jobsuche.

Tritt unserem Telegram-Kanal bei für das, was dir hilft, die Stelle zu bekommen — Gehaltsbenchmarks, den wöchentlichen Marktpuls und neue Feature-Drops. Kein Spam, nur Signal.

Dem Kanal beitreten — kostenlos