Software Engineer II - Platform and Data

About the roleLyraConnect is an AI-native marketing decision platform, purpose built for CPG brands and the retail ecosystem. Our platform runs proprietary demand simulations, targeted and proprietary prompts executed daily across the major AI platforms and AI Assistants, with iterations spanning the occasions and need states that shape a category, revealing how AI responds to shopper need states, which products it recommends, and where a brand is absent from the consideration set. That means we run a lot of prompts against a lot of models and we do the work to turn messy generative output into defensible numbers. This role sits across both halves of that: the backend platform and the pipelines that feed it. You'll own things end to end, ship to production regularly, and have meaningful input at every level of the stack. What you'll work onBuilding and maintaining the LyraConnect PlatformCollection pipelines that query multiple LLM providers at volume, handle rate limits and failures gracefully, and keep results reproducibleParsing and normalizing unstructured model responses into structured brand, product, and sentiment recordsInstrumentation and monitoring so we find out a collection job broke before a client doesChampion our values. Lead in a way that reflects LyraConnect's values framework, I3C3: Imagine, Integrity, Initiative, Collaborate, Communicate, Craft.Qualifications2–5 years of professional experience building web applications, or a CS degree or equivalent coursework, or substantial personal/open-source work you can walk us through2+ years with at least one of: Python, JavaScript/TypeScript, or a comparable languageStrong SQL and a solid understanding of relational database designComfortable on Unix-based systems — you can read logs, trace a process, and debug a server without a GUIStrong Git fundamentalsNice to have but not requiredBS/BA in Computer Science or a related engineering fieldExperience building or maintaining ETL/ELT pipelines and workflow orchestration (Airflow, Dagster, Prefect, or similar)Django or FastAPI service experienceContainerized deployments and functional knowledge of KubernetesWorking knowledge of columnar/analytical databases and dbt-style transformation toolingExperience working with LLM APIs — prompt structuring, output parsing, cost and latency managementApplied statistics background, or enough of one to be dangerousNon-relational databases, caching layers, queueing systemsYou will not have all of these. Apply anyway. How we workNo micromanagement. We hire people we trust to do the right thing and then get out of the way. Learn without permission. Try new approaches. Pick up the tool you need. Nobody has to sign off on you finding a better way to do something. The details matter. Engineering compounds. Small mistakes in a data pipeline become wrong numbers in a client deck three weeks later. Excellence and pragmatism, but not too much of either. Both taken to extremes are failure modes. Knowing when to polish and when to ship is the difference between teams that win and teams that lose. Talk to clients. You should be able to sit in front of a CPG analyst and explain how something works without a translator. LyraConnect, Inc. is an equal opportunity employer.