Senior Data Collection Engineer

Azoma is a hypergrowth startup defining the agentic commerce category. Our revenue has grown tenfold over the past year, and we now work with 20 of the world’s 50 largest CPG companies, alongside other enterprise customers.
We’re building the merchant-side infrastructure for the era of personal shopping agents - transforming how brands and retailers work with AI agents that discover, evaluate and purchase products on behalf of consumers.
Our team combines advanced AI, software engineering and deep commerce expertise. Joining Azoma means taking hands-on ownership as we scale, shaping products used by global brands and helping define how commerce works in the age of AI.
We’re looking for a Senior Data Collection Engineer who enjoys investigating difficult targets, understands why a request or browser session fails, and can turn a successful experiment into a dependable production system.
What you will do
Investigate challenging retailer and conversational-shopping targets; build HTTP and browser collectors that handle dynamic interfaces, session state, and frequently changing response structures.
Turn target-specific discoveries into reusable collection components, diagnostics and operating procedures so each new source does not require another fragile one-off script.
Scale collection through distributed workers, scheduling, queues and concurrency control; tune retries, browser usage and infrastructure providers against throughput and cost per usable result.
Skills and experience
Deep production collection experience on difficult ecommerce or similarly dynamic targets, including incidents where protections or interface changes broke a previously working approach.
Strong programming and asynchronous execution skills in Python, JavaScript or a comparable language, plus practical browser automation with tools such as Playwright or Puppeteer.
An investigative understanding of HTTP, cookies, sessions, proxies, browser fingerprinting and network behavior; you can use captured traffic and browser tools to explain failures.
Experience making collection operate reliably at volume, including distributed queues, retry behavior, monitoring and the tradeoffs between request-based and browser-based extraction.

Work arrangement
Yes