Builder Developer II/III/IV (Senior Agentic Software Engineer)

ECI is hiring senior-levelBuilder Developers(Senior Agentic Software Engineers) focused on rewriting how enterprise software is built, tested, and operated using AI.

The Mission: Use AI to build autonomous agents and workflows that transform legacy ERP systems into modern, AI-driven platforms. You are not maintaining old code, you are building the future. If you enjoy 4-hour "alignment meetings" and writing Jira tickets for things an agent could do in 3 seconds, you’re going to hate it here.

Builderis not just a title; it is a measure ofautonomy, architectural complexity, and "Agentic Leverage."While a traditional "Senior Software Engineer" might be focused on completing tickets, aBuilder is focused on automating workflows and owning the execution of AI-driven modules.

Here are the Builder levels we are hiring:

• BUILDER 2 (Senior): Focuses onAgentic Implementation. This is the "Engine Room" of the team. They don't just use AI; they build the agents. They can take a legacy module and build a LangGraph or CrewAI flow to modernize it with minimal supervision.

• BUILDER 3 (Lead): Focuses onSystemic Orchestration. They design the "Guardrail Layers" and the "Eval Frameworks." They ensure that the agents built by the team are deterministic, cost-effective, and scalable.

• BUILDER 4 (Principal/Architect): Focuses onThe Meta-Platform. They build the internal "Builder Tools" that the other three levels use. They are looking at the 12-month horizon of model capabilities and ensuring the ERP architecture is ready for "Autonomous Operation."

Core DNA (Shared Foundation)

• AI-Native Workflow: Use AI tools to accelerate development, testing, and documentation 10x faster than traditional methods.

• Agentic Implementation: Build, debug, and deploy autonomous agents (LangGraph, AutoGen, Playwright-LLM hybrids) that navigate complex ERP schemas and business logic.

• Legacy-to-Modern Bridge: Treat 20-year-old legacy systems as puzzles for AI-driven modernization.

• Deterministic Mindset: Build for 100% reliability with automated validation, evals, and guardrails.

What You’ll Actually Do

• Build Execution Agents: Move beyond RAG intoAction. Build agents that can navigate an ERP schema, identify a data discrepancy, and suggest a fix.

• Automate the "Untestable": Use AI to wrap legacy business logic in modern unit tests and documentation.

• Prompt Engineering & Evals: Use frameworks likeDSPy to move from "vibes-based" prompting to programmatic, optimized instructions.

• Rapid Prototyping: Turn a "what if" into a working AI-driven feature in days, not months.

The Stack

Python, SQL (Postgres/SQL Server), TypeScript, OpenAI/Anthropic, pgvector/Qdrant, LangChain/LangGraph, dbt, Apache Hop/Airflow, Docker, Weights & Biases, OpenTelemetry. And, any other tool that helps us achieve greatness.

Who You Are

• The "Early Adopter" Archetype: You were using GPT-3 before it was cool and had a local Llama running the week it leaked. You spend more time onr/LocalLLaMA,GitHub Trending, andX (AI Twitter) than on LinkedIn.

• Agent-First Thinker: You don't think in "functions"; you think in "capabilities." You’ve built at least one autonomous agent or complex LLM chain that actuallydoes something useful.

• The "Manual Task" Hater: You have a physical reaction to repetitive work. If you have to do a task twice, you’ve already started drafting the Python script or agentic workflow to automate it.

• Technically Fearless: You aren't intimidated by spaghetti legacy code. You see a 20-year-old undocumented ERP module as a goldmine for AI-driven refactoring and modernization.

• Rigor over Hype: While you love the latest models, you care aboutEvals. You know the difference between a "cool demo" and a "production-grade agent" that handles financial data with 100% accuracy.

• The Tribe Member: You want to work in an environment where your "AI hacks" and custom automation scripts are celebrated as your greatest contribution, not hidden as "side projects."

Why Join Us?

• The "Final Boss" of Complexity: ERPs are the ultimate challenge for AI. If you can automate a General Ledger or a global supply chain, you’ve mastered the hardest problem in software.

• Agentic Autonomy: We don't do "tickets." We do missions. You’ll have the freedom to choose the models, frameworks, and agent architectures you believe will win.

• The BUILDER Tribe: You won't be the "only AI person" in the room. You’ll be part of a 4-person squad of like-minded enthusiasts who are all obsessed with moving 10x faster.

• No Red Tape: If the agent works and the evals pass, we ship. We don't have a "Change Management Committee." We have a "Does It Work?" policy.

• The Final Boss: ERP data is the hardest problem in software. If you want an easy life, go build another photo-sharing app. If you want to win the game, let’s go!