AI Integration for Salesforce File Download

Salesforce + OpenAI API Automation for CV Review and Candidate Summaries Project Overview We operate a recruitment platform built on Salesforce.com and want to automate an existing manual process that uses ChatGPT to review candidate CVs against individual job requirements. We are looking for an experienced Salesforce developer/integration specialist who can design, build, test and deploy this automation. Current Salesforce Structure Our recruitment platform has three key Salesforce record types/objects: Contacts – candidate/person records. Job Orders – individual vacancies/roles being recruited. Job Applications – records linking a Contact/candidate to a Job Order. When somebody applies for a role through our online job board: A Contact record is created or updated for the candidate. Their CV/resume is stored in Salesforce as a file/attachment. A Job Application record is created linking the candidate to the relevant Job Order. The Job Order contains a prompt/instruction that we currently use when reviewing candidates with ChatGPT. Current Manual Process At present our team manually: Opens the candidate's Salesforce Contact/Application. Downloads or copies the candidate's attached CV. Provides the CV to ChatGPT. Provides ChatGPT with a review prompt stored on the relevant Job Order. Asks ChatGPT to review the CV against the requirements of that role. Receives a candidate summary/review from ChatGPT. Copies the resulting summary back into a field on the candidate's Job Application record. We want to automate this end-to-end. Required Automation When a new Job Application is created, or when an appropriate trigger/condition is met, the system should: Identify the associated Contact/candidate. Retrieve the candidate's current CV file from Salesforce. Retrieve the AI review prompt/instructions from the associated Job Order. Submit the CV content and Job Order prompt to the OpenAI API. Receive a structured candidate assessment/summary. Write the resulting summary into the appropriate field on the Job Application record. Record processing status and any errors so failed assessments can be identified and re-run. The solution should be reliable, secure and suitable for production use. Important Requirements The contractor should be comfortable with: Salesforce architecture and development Salesforce Files / ContentDocument / ContentVersion or legacy attachments where applicable Salesforce Flow and/or Apex Salesforce REST/API integrations OpenAI API integration Secure storage of API credentials Handling PDF, DOC and DOCX CV/resume files Extracting readable text from CV files Triggering automated processing from Salesforce Updating Salesforce records after API responses Error handling, logging and retry mechanisms Salesforce governor limits and asynchronous processing Testing in a Salesforce sandbox before production deployment We would like the architecture to avoid unnecessary manual intervention and to be maintainable by another Salesforce developer in future. AI Output The prompt will vary depending on the Job Order. The OpenAI response may include information such as: Candidate overview Relevant experience Alignment with the role Relevant skills Potential concerns or gaps Overall suitability Recommended next step The exact format will be agreed during the project. Where practical, we would prefer the OpenAI response to use a consistent structured format rather than completely free-form text. Security and Privacy Candidate CVs contain personal information, so the contractor must take data handling and security seriously. The solution should: Send only information required for the assessment. Secure API credentials appropriately. Avoid unnecessarily storing candidate information outside Salesforce. Document where information is transmitted and retained. Follow Salesforce and OpenAI recommended security practices. Project Stages We expect the project to be completed approximately as follows. Stage 1 – Discovery and Solution Design Review our existing Salesforce setup and confirm: Salesforce objects and fields involved File storage method Trigger point for the automation CV text extraction method OpenAI API architecture Security/authentication approach Error handling Estimated API usage/cost Recommended Salesforce architecture Provide a short written implementation plan before development begins. Stage 2 – Prototype Build a working prototype in our Salesforce sandbox that can: Read one Job Application Retrieve its Job Order prompt Retrieve the candidate CV Send the required information to OpenAI Receive a response Save the response to the Job Application Stage 3 – Production-Ready Automation Add: Automatic triggering Processing status fields Error logging Retry/reprocess capability Protection against duplicate processing Appropriate asynchronous processing Testing for common CV file formats Appropriate OpenAI model/configuration settings Stage 4 – Testing and Deployment Test multiple applications and job orders. Resolve errors and edge cases. Deploy to production. Verify production operation. Stage 5 – Documentation and Handover Provide documentation covering: System architecture Salesforce fields/components created OpenAI configuration API credential configuration Trigger logic Error/retry process How prompts can be changed How the integration can be maintained How the OpenAI model/configuration can be changed in future We would also like a short handover session with our team. Deliverables The successful contractor will deliver: Approved solution design Working Salesforce/OpenAI integration Sandbox implementation and testing Production deployment Error logging/reprocessing capability Source code/configuration Technical documentation Administrator/user documentation Handover session Skills We Are Looking For Please apply only if you have strong experience with at least several of the following: Salesforce Development Apex Salesforce Flow Salesforce REST integrations Salesforce Files OpenAI API AI/LLM integrations PDF/DOCX text extraction Recruitment or ATS systems API security and authentication Experience integrating OpenAI or another LLM directly into Salesforce would be particularly valuable. When Applying Please answer the following: Describe a Salesforce/API integration you have personally built. Have you integrated Salesforce with OpenAI or another LLM? If yes, briefly describe it. How would you recommend triggering this process when a new Job Application is created? How would you extract text from PDF and DOCX CVs stored in Salesforce? Would you use Salesforce Flow, Apex, middleware, or a combination? Explain why. How would you handle an OpenAI API failure or a CV that cannot be processed? How would you prevent the same Job Application from accidentally being processed multiple times? How would you securely manage the OpenAI API credentials? What information would you need from us before providing a fixed estimate? Please confirm that you can provide documentation and hand over the completed solution to another Salesforce developer. Please do not submit a generic AI-generated proposal. We are particularly interested in your proposed technical approach and examples of comparable work you have personally completed.