We've got around 20 Java / Spring Boot / Apache Camel services that basically connect everything at PopSockets that needs to talk to something else. NAV and Business Central on the ERP side, a 3PL that ships every B2B and B2C order, Salesforce Commerce Cloud, EDI with the big retailers (Target, Amazon, Best Buy, etc), a fulfillment partner in China, our PLM, and a data platform that wants a copy of all of it. It all runs on Azure Kubernetes. We're looking for a senior dev who can own a chunk of that. Honestly, adding bodies doesn't add much shipping capacity unless the dev is legitimately good. Building up all the context and requirements somebody needs to do the job well takes real time, so we're being picky. The person we want is the one who can pick up "orders stopped showing up at the warehouse this morning," run it to ground across three systems, fix it properly, and write down what happened so the next person doesn't have to. The team is split between India and Colorado and you'd be working day to day with devs and ops folks on both sides. There's real overlap required, more on that below. We've got around 20 Java / Spring Boot / Apache Camel services that basically connect everything at PopSockets that needs to talk to something else. NAV and Business Central on the ERP side, a 3PL that ships every B2B and B2C order, Salesforce Commerce Cloud, EDI with the big retailers (Target, Amazon, Best Buy, etc), a fulfillment partner in China, our PLM, and a data platform that wants a copy of all of it. It all runs on Azure Kubernetes. We're looking for a senior dev who can own a chunk of that. Honestly, adding bodies doesn't add much shipping capacity unless the dev is legitimately good. Building up all the context and requirements somebody needs to do the job well takes real time, so we're being picky. The person we want is the one who can pick up "orders stopped showing up at the warehouse this morning," run it to ground across three systems, fix it properly, and write down what happened so the next person doesn't have to. The team is split between India and Colorado and you'd be working day to day with devs and ops folks on both sides. There's real overlap required, more on that below. We build with AI. A lot. Read this part twice. AI coding agents aren't a side tool here, they're how the work actually gets done. We use them to write code, review PRs, dig through logs, draft runbooks and work production issues. Every repo has agent instructions checked in and an AI review runs on every PR before a human looks at it. We expect you to work this way from day one and get better at it every month. That does not mean "let the model write it and ship it." It means you know how to drive an agent, check its work, spot when it's confidently wrong, and set it up so it's actually useful on our codebase instead of generic. If you're skeptical of the whole thing, or you've only ever used autocomplete, this probably isn't the right fit. If you've already rebuilt your workflow around agents and you're a little annoyed at teams that haven't, let's talk. What you'd actually be doing - Build and maintain Camel routes and Spring Boot services. New integrations, changes to existing ones, the occasional "why is this taking 40 seconds" rabbit hole. - Own prod. When an order doesn't reach the warehouse or a shipping confirmation doesn't come back, you're one of the people who figures out why and fixes it. Some of that is code, a lot of it is reading logs and data and asking the right person the right question. - EDI. 850 / 940 / 945 / 856 / 810 with SPS Commerce and direct retailer connections. If you haven't done EDI before you can learn it, but you have to actually want to. - Work with the ERP and ops teams. Half the bugs in an integration platform are really data or process problems upstream. The good engineers here are the ones who can tell the difference and say so without being a jerk about it. - Ship through our pipeline: GitHub Actions, Helm, AKS, Spring Cloud Config + Azure Key Vault, Azure Service Bus, Postgres, Datadog. You don't need to be an infra person but you need to be comfortable reading a Helm chart and a pod log. - Work agent-first. Use AI to write and review code, run investigations and produce docs, and help keep our agent instructions and tooling sharp so everybody benefits. - Review PRs, get yours reviewed, leave the codebase a little better than you found it. We care about fail-fast config, useful logging, and not building things that need a human to babysit them. We also care about not over-engineering. Simple and boring wins. Operating Model - Individual contributor on the Integrations team, reporting into US engineering leadership. Works closely with the India ERP and data teams. - Timezone. This isn't a follow-the-sun role and it isn't a US-hours role either. What we need is reliable overlap with Mountain time in the US morning, which is your evening. Practically that means being genuinely available roughly 7 to 10 pm IST most days for standups, pairing and prod issues, and the rest of your day is on your own schedule. The whole point is to avoid the 24-hour response cycle where somebody hits a wall, posts about it, and then waits a full day for an answer. When something's on fire in prod we'll need you outside those hours sometimes, and we'll be flexible back. What we're looking for - Immediate joiner who can start in Oct or early Nov - 6+ years building backend Java, at least a few of them on Apache Camel. We're on Camel 4 / Spring Boot 3 / Java 21, so recent experience matters more than ancient versions. - Daily, hands-on use of AI coding agents (Claude Code, Cursor, Copilot agent mode, or similar) as your primary way of writing and reviewing code. Not "I've tried ChatGPT." You should be able to talk about what works, what doesn't, and how you keep it honest. - Real experience running integration services in prod: queues, retries, idempotency, dead-letter handling. The stuff that separates "works on my machine" from "works at 2 am." - Comfortable in SQL and Postgres. You'll spend a lot of time querying tables to figure out what actually happened. - Kubernetes and a cloud provider. We're Azure but AWS or GCP translates fine. Helm, containers, reading logs and metrics. - Someone who writes things down. Runbooks, PR descriptions, the note in the ticket explaining the fix. We're spread across two continents and the doc is how the other side finds out. - Surfaces blockers the day you hit them, not three days later in a status update. If you're stuck, say so while somebody can still help. - Can work independently across a time gap. Knows when a decision needs the other side of the world awake and when it doesn't. Nice to have - EDI, especially retail (SPS Commerce, X12 850/856/945). Big plus. - Microsoft Dynamics NAV / Business Central, or any ERP with a real order-to-cash flow. - 3PL / warehouse integrations, order management systems. - Salesforce Commerce Cloud, Shopify or marketplace APIs. - Spring Cloud Config, Azure Service Bus, APIM, Datadog. - Snowflake, ADLS, Microsoft Fabric, or otherwise feeding a data platform from operational systems. PopSockets is dedicated to the practice of equal opportunity employment. We prohibit unlawful discrimination against applicants and employees on the basis of age, race, sex, sexual orientation, gender identity, religion, national origin, disability, military status, genetic information, color, creed, ancestry, or any other status protected by applicable federal, state or local law. This prohibition includes unlawful harassment based on any of these protected classes. Unlawful harassment includes verbal or physical conduct which has the purpose or effect of interfering with an individual's work performance, or creating an intimidating, hostile, offensive, unsafe or otherwise non-welcoming work environment. This policy applies to all employees, including managers, supervisors, co-workers; and non-employees such as customers, clients, vendors, consultants, etc.

6-months Contractor - Senior Integration Developer (Apache Camel / Java)

similar openings