A & R Autoparts

Senior Full-Stack Developer

Job Summary

  • Contract
  • Remote

Job Overview

About the role

**A & R Autoparts Senior Full-Stack Developer — Custom CRM, Web**

Apps & AI Agents

Experience: 5–7 years | Type: Contract (milestone-based), with potential for long-term engagement | Location: Remote

Who We're Looking For

We're hiring a senior developer who has built custom CRMs, web applications and AI agents from scratch, and has done it before in production. This is not a role for someone who has only configured off-the-shelf tools or maintained someone else's system. You should be comfortable owning a product from database design to deployment, and be familiar with our current CRM stack (below). Experience with Node.js and Python for machine learning is required.

Our Current Stack

Laravel (PHP) with MySQL 8 Inertia.js with React RingCentral (voice & SMS), Microsoft 365 / Outlook (email), Google Ads (lead generation) Shopify / Stripe (payments)

What You'll Do

Custom CRM development Own the architecture and roadmap of our in-house CRM: lead, contact, deal, quote and task modules Upgrade and modernize the codebase (current PHP and Laravel versions, updated front-end packages) with no loss of features Design clean MySQL schemas, migrations, role-based access and audit logging

Build dashboards and reporting for sales performance, lead sources and conversion Web applications Build new internal and customer-facing web apps and services around the CRM Develop Node.js services where they fit best (real-time features, integrations, background workers) Write clean, tested, well-documented code with a solid deployment process AI agents & machine learning

Build AI agents that act inside the CRM, for example

Call transcription and extraction of vehicle details (year, make, model, engine, VIN) into structured CRM fields A sales-assist agent that drafts quotes, suggests follow-ups and summarizes customer history An inbox/SMS triage agent that classifies inbound messages and routes or drafts replies Use LLM APIs (Claude, OpenAI) with tool/function calling, structured JSON output, prompt design and evaluation Build Python ML models where they add value, such as lead scoring, pricing suggestions, or call-outcome prediction Add guardrails: human-approval steps, logging, cost controls and fallback behavior when the model is unsure Automations & integrations Integrate RingCentral (webhooks, call events, SMS, softphone embedding): every inbound call creates or matches a lead; missed calls trigger text-back and callback tasks Build automated follow-up sequences with queues and the scheduler Send quotes by SMS and email (Microsoft Graph) with payment links (Stripe or Shopify) and handle payment webhooks Upload offline/call conversions to Google Ads Connect third-party tools with APIs, webhooks, or platforms like n8n, Make or Zapier where it makes sense

Must-Have Skills

5–7 years of professional software development

Proven experience building custom CRMs, web apps and AI agents from scratch (portfolio or project examples required) Strong Laravel/PHP, including upgrading legacy apps across major versions React with Inertia.js MySQL schema design, indexing and migrations Node.js for backend services, APIs and real-time features Python for machine learning (e.g. scikit-learn, pandas, PyTorch or TensorFlow) Production experience with AI agents / LLM integrations: tool calling, structured outputs, prompt engineering, and handling unreliable model output Laravel queues (Redis/Horizon), task scheduler, and broadcasting (Pusher, Reverb or similar) Reliable webhook design: idempotency, retries, dead-letter / failed-job handling Hands-on telephony or SMS API work (RingCentral preferred; Twilio or Vonage fine) Clear written English and regular, proactive progress updates

Nice to Have

RingCentral API specifically Speech-to-text APIs (Whisper, Deepgram, AssemblyAI, etc.) Agent frameworks or protocols (MCP, LangChain/LangGraph, Claude Agent SDK,

OpenAI Agents SDK)

RAG / vector search (pgvector, Pinecone, Qdrant) for knowledge-based agents Workflow tools: n8n, Make, Zapier Stripe, Shopify, Microsoft Graph and Google Ads APIs Familiarity with A2P 10DLC and TCPA rules for SMS Experience in automotive, parts, or phone-sales businesses

How We'll Work

Paid per milestone, each with clear acceptance criteria All code goes into our Git repository; all API and AI accounts stay in our company's name Documentation and a handover session at the end of each phase We start with a paid 1-week test task: in a RingCentral sandbox, an inbound call creates a lead in a Laravel app, a missed call sends an automatic text, and an LLM extracts vehicle details from a sample call transcript into structured fields