Automation Engineer

WHAT THIS ROLE ACTUALLY IS

You eliminate manual work. Every repetitive process a human is doing by hand across this company's commercial operations, government contracting workflows, and marketplace channels is a target for automation. You identify it, scope it, build the automation, test it against every edge case, deploy it, and monitor it. You do not build automations that require human babysitting after deployment. You build automations that run unattended and alert only when something genuinely needs human judgment. You use AI and LLMs as part of your automation stack — wherever an LLM makes the automation more capable than a rule-based system, you use it. Wherever a rule-based system is more reliable and auditable, you use that instead. You make the right technical decision for the specific task, not the most impressive-sounding one.

WHAT YOU DO EVERY SINGLE DAY

• Identify automation opportunities across every department proactively — you do not wait for a request. You analyze workflows, identify manual steps, calculate time cost, and propose automations with a documented ROI estimate
• Build no-code and low-code automations using Make.com, Zapier, or n8n — connecting any two systems with an API faster than a developer could write the code from scratch
• Build coded automations in Python when the task requires custom logic, complex data transformation, or performance that no-code tools cannot deliver
• Integrate LLMs into automation workflows for document processing, data extraction, classification, content generation, and routing decisions — using AI where it makes the automation genuinely more capable
• Build government contracting automations — SAM.gov monitoring, opportunity scoring, proposal section routing, compliance checklist generation, pricing data extraction from USASpending.gov
• Build marketplace automations — Amazon listing updates, inventory alerts, pricing rule enforcement, sales data aggregation across Shopify and Walmart, order routing
• Build AI-powered document processing pipelines — intake an RFP, extract all requirements, classify by section, map to capabilities, output a structured brief without human intervention
• Test every automation against edge cases before deployment — malformed input, API unavailability, rate limit hits, unexpected LLM output format — with an answer for every scenario
• Monitor all deployed automations — you know which ran last night, which failed, and which produced outputs needing human review before anyone tells you
• Document every automation — inputs, outputs, triggers, failure handling, dependencies, and how to disable it if something goes wrong
• Submit a written daily standup update every working day without exception

TECHNICAL REQUIREMENTS — CORRECTED FOR GRADE LEVEL I (3–5 YEARS MINIMUM)

Automation Platforms (4 yrs total / 2 yrs complex production workflows): Minimum 4 years of professional experience building production automations using Make.com, Zapier, n8n, or equivalent — of which at least 2 years must be complex multi-step workflows with conditional branching, error handling, retry logic, and data transformation running unattended in production. You have maintained automations you did not build and you have fixed automations that failed at midnight. You know which platform breaks under which conditions.

Python (3 yrs professional automation): Minimum 3 years of professional Python for automation, data processing, and API integration. You write clean, error-handled, documented Python that a colleague can maintain without calling you. You have built Python automation scripts that have run in production for over 12 months without requiring your intervention.

API Integration (4 yrs REST / 2 yrs webhooks): Minimum 4 years of professional experience consuming REST APIs across a wide variety of enterprise platforms — every authentication pattern, pagination scheme, rate limit strategy, and error response format. You do not need documentation for common platforms. Minimum 2 years of webhook configuration and reception — including security implications of public webhook endpoints and validation implementation.

LLM APIs and AI Integration (2 yrs production pipeline automation): Minimum 2 years of professional experience integrating LLM APIs into automated production workflows — not chatbot integration but pipeline automation. Document processing, data extraction, classification, and routing where LLM output feeds downstream automation steps without human intervention. You have a documented strategy for when the LLM does not return the expected output format.

Data Handling (3 yrs professional / 1 yr SQL): Minimum 3 years of professional experience transforming and routing data between systems — JSON, CSV, XML, and structured database formats. You have built data pipelines that process high volumes without losing records or introducing duplicates. Minimum 1 year of SQL querying for data extraction and aggregation.

Monitoring and Observability (3 yrs production): Minimum 3 years of building and maintaining monitoring systems for production automations — error logging, execution tracking, failure notification, and performance monitoring. You have configured alerting that reaches you before the affected team. You built the monitoring that told you.

Government and Enterprise Data (2 yrs regulated environment): Minimum 2 years of building automations handling regulated data — CUI, PII, HIPAA, or equivalent. You have designed audit trail logging that satisfies a compliance requirement. You know which data cannot touch which systems and you enforce it in the automation design.

WHAT WE REQUIRE — NO EXCEPTIONS

• 3–5 years of professional automation and integration experience with at least 2 years of complex production automations running unattended
• You have integrated an LLM into an automation workflow — not a chatbot, an automated pipeline. You can describe the specific task, the model, and how you handled unexpected LLM output
• You have used Make.com, Zapier, n8n, or equivalent at a professional level — complex workflows with conditional logic and production-grade error handling
• You can write Python well enough to build a data processing script, call an API, handle exceptions, and log activity — without requiring a software engineer to implement your ideas
• PST Time Zone
• Professional written English — documentation, daily updates, and task communication are clear and complete
• If an automation fails at night, you have identified it and started the fix before anyone asks