BUILDER 3 Systemic OrchestrationLevel 3 Lead Technical Direction The Squad You are the technical lead of the Modernization Strike Team a small squad of Builders rewriting how enterprise software is built tested and operated using AI You are not the people manager You are the person whose judgment the squad borrows when the agent works on Tuesday and fails on Friday The Mission A Builder 2 makes an agent work You make it work every time at a cost we can defend on systems we cannot afford to break Your job is to turn a collection of clever agents into a dependable production capability across a 20 year old ERP estate If you think it worked in the demo is a finish line you will hate it here The Builder Levels A Hierarchy of AccelerationBUILDER 1 Professional AI Assisted Execution Uses AI to write code tests and documentation 3x faster than a traditional junior Learning to string together basic chains and promptsBUILDER 2 Senior Agentic Implementation The engine room Doesnt just use AI; builds the agents Takes a legacy module and builds a LangGraph or CrewAI flow to modernize it with minimal supervisionBUILDER 3 Lead Systemic Orchestration You are here Designs the Guardrail Layers and the Eval Frameworks Ensures the agents built by the squad are reliable cost effective and scalableBUILDER 4 PrincipalArchitect The Meta Platform Builds the internal Builder Tools the other three levels use Works the 12 month horizon of model capability and readies the ERP architecture for autonomous operationThe honest boundary B2 owns an agent B3 owns whether the squads agents are allowed near production B4 owns the platform the squad builds on If you are still measured mainly by the modules you personally shipped you are a strong B2 If you are measured by whether other peoples agents survive contact with production you are a B3 The Builder 3 ProfileYou are a systems engineer for non deterministic systems You are comfortable saying no not yet to a working demo and you can say exactly what would change your mind 1 Core DNA shared foundation at lead altitudeAI Native Workflow You dont just use AI tools fast you set the squads bar for what fast and secure looks like and you remove the bottleneck when someone hits oneAgentic Architecture You still build end to end agents yourself and you do But your leverage comes from designing the state management tool interfaces and failure semantics that everyone else builds insideLegacy to Modern Sequencing You decide which legacy modules get modernized first based on blast radius data gravity and how cheaply success can be measured not on which one looks most funMeasured Reliability You define what reliable enough means for each workflow acceptance thresholds regression suites human in the loop checkpoints rollback paths You know that 100 accurate is a marketing phrase and that the engineering version is a measured error rate with a defined containment strategy2 The Three Things You Owna The Guardrail Layer Input and output validation tool permission scoping financial and PII data handling sandboxed execution human approval gates on irreversible actions kill switches blast radius containment You assume the agent will eventually do the wrong thing and you design so that it costs us a log line instead of a General Ledger b The Eval Framework Golden datasets built from real ERP edge cases Offline evals in CI as a merge gate online evals in production LLM as judge calibrated against human labels and you know when that calibration has drifted Regression detection when a model version prompt or retrieval index changes underneath you You move the squad from prompt vibes to a number that can fail a build c The Cost &

- Performance Envelope Model routing caching context budgeting latency budgets You track token cost per completed business transaction not per call and you can tell the CFO why a workflow costs what it costs and where the next 40 comes from 3 Elective Depth be deep in one fluent in all sixAgentic Architect Systems state management and tool calling interfaces for long running business processes across complex ERP schemasUXInteraction Product generative UIs and browser use agents; designing where the human stays in the loop and where they dont need to beData &
- Context Data RAG and knowledge graph architecture vector stores chunking strategy retrieval evals and the relationship between a Purchase Order and a GL entryReliability &
- Eval QA programmatic prompt optimization DSPy red teaming agents adversarial testing of financial logicInfra &
- Agent Ops DevOps the factory where agents live tracing logging and monitoring the thoughts of the AI; deployment and rollback of prompts and models as versioned artifactsDomain Logic Translation Business turning 50 pages of accounting regulation into structured system prompts and agentic constraintsWhat Youll Actually DoSet the definition of shippable Write the standard for when an agent can touch production data and enforce it in code not in a wiki pageBuild the eval harness and wire it into CI as a gate that can and does block mergesDesign the guardrails around regulated and financial logic including the human approval and rollback pathsRun design reviews Kill bad agent architectures in week one instead of month four Be specific about whyOwn agent incidents Pull the traces find the root cause and add the eval that makes that failure impossible to reintroduceMake build vs buy calls on frameworks vendors and models and revisit them when the ground shifts which it willGrow Builder 1s and 2s through code review design review and pairing Not through ceremonyWhat Good Looks LikeTargets to be agreed with the hiring manager these are illustrative not fixed 90 days Eval harness running in CI on at least two live agentic workflows with a published reliability baseline A written enforced guardrail standard6 months Every production agent has a regression suite a cost per transaction number and a rollback path Agent related incidents are down and post incident evals exist for each12 months The squad ships agentic modules on a predictable cadence without you in the critical path of each oneThe StackPython SQL PostgresSQL Server TypeScript OpenAIAnthropic pgvectorQdrant LangChainLangGraph DSPy dbt Apache HopAirflow Docker Weights &
- Biases OpenTelemetry plus the orchestration layer youll live in eval and tracing platforms LangSmith Langfuse Braintrust or equivalent policy and guardrail tooling model routing and caching And any other tool that helps us achieve greatness Who You AreYouve been burned You have shipped an agentic system to real users watched it fail in a way you didnt predict and changed your architecture because of it This is the single strongest signal for this levelRigor over hype You love new models and you still ask for the eval numbers first You know the difference between a cool demo and a production grade agent handling financial dataAgent first thinker You think in capabilities and failure modes not functionsThe manual task hater If you do a task twice youve already started drafting the workflow that kills itTechnically fearless A 500 table undocumented legacy schema is a puzzle not a threatA multiplier Your best week is one where the squad shipped three things and you wrote one design doc and two brutal useful reviewsWhy Join UsThe Final Boss of complexity ERPs are the hardest problem in software for AI Automate a General Ledger and youve mastered the gameAgentic autonomy We dont do tickets We do missions You choose the models frameworks and architectures you believe will winNo red tape If the agent works and the evals pass we ship No Change Management Committee a Does It Work policy Note that you are the person who defines what works meansThe Builder Tribe You will not be the only AI person in the room Youll lead a squad of people as obsessed with moving 10x faster as you are

📌 Builder 3 (India)
🏢 Eci Software Solutions
📍 India