> Markdown version of [/jobs/ext/4544327-ai-automation-crm-integration-engineer](https://www.wearedevelopers.com/jobs/ext/4544327-ai-automation-crm-integration-engineer). Every page supports `.md` or `Accept: text/markdown`. Links point to the HTML versions so they work for humans too. Agent guide: [/agents.md](https://www.wearedevelopers.com/agents.md). --- # AI Automation & CRM Integration Engineer - **Company:** CentralReach, LLC - **Location:** Holmdel, United States - **Salary:** $100,000 - $125,000 - **Contract:** Permanent contract - **Skills:** Application Programming Interfaces (APIs), Artificial Intelligence, Airflow, Extract Transform Load (ETL), Data Systems, Python (Programming Language), OpenAI, Standard Sql, Salesforce.Com, Tableau (Software), Workflow Management Systems, AI Infrastructure, Pinecone, Ontology Design, Retrieval-Augmented Generation, Large Language Models, Prompt Engineering, Llamaindex, Git, HuggingFace, RAGAS (Retrieval Augmented Generation Assessment), Data Analytics, Weaviate, Nintex, Machine Learning Operations, FAISS, Evaluation of Large Language Models, Restful APIs, Semantic Modeling, Model Inference, Data Pipelines, Automation Anywhere - **Published:** November 3, 2026 - **Apply:** https://startup.jobs/business-systems-automation-engineer-centralreach-company-10653877 ## About the Role We value demonstrated ability over credentials. If your work shows the thinking, we want to hear from you. ## Description * Design and deploy multi-step AI automation flows that connect LLM inference, CRM triggers, webhook events, and data transforms across integrated systems * Build and maintain full-stack integrations between large language model APIs and Salesforce - reading and writing CRM context dynamically into model prompts and back into SFDC objects and flows * Engineer prompt templates, few-shot frameworks, and output validation logic to ensure AI responses meet business-defined quality standards * Develop and run LLM evaluation pipelines using frameworks such as RAGAS, TruLens, or equivalent, measuring model performance against real business outcomes * Collaborate with Tableau and Salesforce stakeholders to surface AI-driven insights through dashboards and reporting workflows * Translate technical model outputs into interpretable business signals, supporting revenue operations and pipeline visibility Desired Skills and Experience: Required: * Multi-step AI automation Builds end-to-end AI workflows across connected systems - triggering model calls, routing outputs, and updating downstream platforms without manual handoffs. * Full-stack LLM + Salesforce integration Connects LLM APIs directly to Salesforce objects and flows - designing pipelines that pull live CRM context into model prompts and write structured outputs back into SFDC. * Prompt engineering & LLM evaluation Engineers prompts systematically and measures their performance - from few-shot design and output validation to RAG retrieval quality and business-aligned evaluation frameworks. * Knowledge-based AI & ontology design Understands how to structure what a system needs to know - defining entity types, relationships, and semantic hierarchies so that AI pipelines reason accurately over real-world information. Preferred: * Exposure to revenue operations workflows, pipeline reporting, or CRM-driven business processes * Experience working with Tableau or SFDC dashboards in a cross-functional analytics context * Ability to frame AI outputs in terms of business value - lead conversion, churn reduction, process efficiency Tech stack signals we look for Python · SQL · LLM APIs (OpenAI / Anthropic / HuggingFace) · Salesforce (RESTAPI / Flows / Objects) · RAG frameworks (LangChain, LlamaIndex) · Vector databases (FAISS, Pinecone, Weaviate) · Ontology / semantic modeling (RDF, OWL, SKOS or equivalent) · Workflow orchestration platforms (N8N, Airflow, or equivalent) · MLOps / managed AI infrastructure (Shakudo, SageMaker, or equivalent) · Tableau · ETL / data pipeline tooling · Git You might be a fit if you've * Built a chatbot or assistant that reads from and writes to a live data system * Designed a RAG pipeline where the retrieval layer was structured around real business entities (accounts, contacts, opportunities) * Written evaluation logic that tests model output quality - not just accuracy, but relevance and business alignment * Modeled an information flow - even in a project setting - where entity relationships shaped how a system reasoned or retrieved