About Us Paragon International, Inc. is a U.S.-based manufacturer of commercial concession equipment and food service products. We receive purchase orders from customers such as Amazon, Home Depot, distributors, school systems, and other commercial customers. Orders arrive by email in many different formats, including PDFs, Word documents, Excel spreadsheets, scanned documents, and occasionally photographed purchase orders. We are looking for an experienced AI Automation Engineer to design and build a production-ready system that automates our entire order intake process. This is not a simple chatbot project. We need someone who has successfully built business automation systems that combine AI, OCR, document processing, APIs, and workflow automation. Project Overview The system will monitor one or more Gmail inboxes continuously and automatically process incoming emails and attachments. The workflow should: * Monitor Gmail 24/7 for new incoming emails. * Download all attachments automatically. * Read: * PDF files * Microsoft Word documents * Excel spreadsheets * Scanned PDFs * Image files (JPG, PNG, TIFF, etc.) * Photographs of purchase orders * Use OCR when required. * Use AI to determine whether the email is: * Purchase Order * Quote Request * Cancellation * Return/RMA * Customer Inquiry * Other * Identify the customer automatically. * Extract all order information into a standardized data structure. * Detect duplicate purchase orders. * Automatically print valid purchase orders to our network printer. * Save documents into organized folders. * Rename files using a consistent naming convention. * Move processed emails into Gmail folders/labels. * Generate logs for auditing and troubleshooting. ## Future Phases The initial project focuses on reliable document processing and printing. Additional phases may include: * Sage 100 ERP integration * Automatic sales order creation * Inventory verification * Customer acknowledgment emails * Shipping workflow automation * Dashboard and reporting * AI exception handling * Multi-location printing We are looking for a long-term development partner who can continue improving the system over time. ## Required Skills Please apply only if you have strong experience with most of the following: * OpenAI API / ChatGPT API * Gmail API * OCR technologies (Tesseract, Azure Document Intelligence, Google Vision, AWS Textract, or similar) * Intelligent Document Processing (IDP) * PDF parsing * Workflow automation * Python * REST APIs * Windows automation * Network printing * Error handling and logging * AI document classification Experience with the following is a significant advantage: * n8n * Microsoft Power Automate * Make.com * ERP integrations * Sage 100 * Purchase Order processing * Manufacturing or distribution businesses ## Deliverables The completed solution should: * Run continuously with minimal supervision. * Be reliable enough for production use. * Handle errors gracefully. * Be well documented. * Be easy for our staff to maintain. * Be scalable as our order volume grows. ## To Apply Please include: 1. A description of similar automation projects you have completed. 2. Which automation platform you recommend (Python, n8n, Power Automate, Make, or another solution) and why. 3. Examples of AI document processing or OCR projects you've built. 4. Your experience integrating with ERP systems. 5. Your estimated timeline. 6. Your hourly rate or fixed-price proposal. Please begin your proposal with the phrase: **"I have built AI document automation systems."** This helps us identify applicants who have carefully read the project description. We are looking for a long-term partner, not just someone to complete a single project. If this project is successful, additional work will include ERP integration, warehouse automation, customer service automation, purchasing automation, and AI-driven business process improvements.

I am working through a design agency on an application for their end client. I think the agency will need you to contract with them directly, but I will manage the project for them. I have scoped out the project already, and our plan is to internally perform a design phase with the client to produce a prototype with Lovable. There may be minor changes to scope after that design phase. The purpose of the app is to create bespoke wedding gown concept images for potential customers of an online wedding dress store. I have provided the details below and attached as a PDF 1. Customer opens an AI dress/gown design experience from a link in their separate e-commerce site. - This can be presented in its own page, we don't want a chat window to be present on any other page - This will be a chat-based interface built into the content area of the page, instead of a popup - The design must be elegant, and match the theme of the e-commerce site - The top navbar and footer don't need to be exactly recreated in this subdomain site, but should look similar enough to create a seamless experience - There will be no integration with the e-commerce site, we need to keep these web apps completely separate 2. The customer must sign up for an account and purchase one credit to begin the AI session - We will need to set up the subdomain site with its own payment processing system and login system - The payment integrations are Stripe to facilitate credit card and Apple pay, and a basic Paypal integration 3. The customers should be able to use a magic link to sign into their accounts, instead of having to remember a password - The account should automatically remember the browser to reduce friction for future access to the app 4. When an AI session begins, we will ask the customer a series of questions programmatically to prime the AI agent so that it can deliver better results - The questions will need to use conditional logic, such that the first question which determines one of 3 main conditional tracks: What type of gown are you looking for? Wedding Gown, Evening Gown, Cocktail Dress - If Wedding Gown is selected, the AI should suggest for the customer to take go to a bridal store and pictures of themselves in different dresses they like and upload the pictures, and describe what they do and don’t like about each dress. It can ask this in the freeform chat, since it may make the most sense to let them fill out the entire questionnaire to stay engaged, and we should reduce the costs of development for the questionnaire by omitting any unnecessary UI that the freeform chat can provide. - It may be best to always just prompt for them to upload the inspirational image at the beginning of the freeform chat so we can omit unnecessary programmatic UI, but in the case of the Wedding Gown it will specifically ask the customer to peform the above task. - We may have other specific questions to add to the questionnaire depending on what conditional track the customer chooses, though only the 3 main branches of conditional logic based on dress type will be required. - Examples of general questions it will need to ask are as follows: -- silhouette -- neckline -- sleeves -- fabric -- embellishments -- color -- train -- length -- closure -- lining -- structure -- inspiration -- event type -- I didn’t get the exact list of questions yet from my client that we should ask in the initial questionnaire. Let me know if you will need this information to accurately provide a price for the development of this application 5. We should not display a concept image after the programmatic questionnaire, the customer will be taken directly into the freeform chat from the questionnaire. - The AI agent may start with an overview of the selected choices from the questionnaire, then can generate concept images at its discretion. 6. The AI should guide the customer through a freeform conversation - The conversation should begin with the AI asking the customer to subjectively describe their dream dress 7. The AI should also make a suggestion near the beginning of the conversation for the customer to upload at least one inspirational photo, but photo upload is optional - If the customer uploads an initial inspiration image, the AI agent should not attempt to figure out body type, measurements, or any other information that we can gather programmatically. - It should treat the inspirational image the same way it would treat any image the customer uploads during the freeform chat, to reduce the cost of development as much as possible. 8. Customer can proceed with a freeform conversation description - The customer should have the option to type in a chat and to upload images - The purpose of the conversation is for the customer to describe the desired dress or gown in an open-ended way 9. AI generates one or more concept images based on the conversation, as soon as it can once it has enough information - The AI model we select should be very good at generating these types of images, this is probably the most important quality the AI model needs to have - The concept images should have the same quality as the final image 10. It is acceptable to generate the gown on a mannequin or a real human model, however the dress must be photorealistic, not a sketch or cartoonish rendering. 11. The concept images that the AI generates and the final image should portray the body type and skin color which the customer specifies - It is very important for us to render the image of the garment on the correct body type - ex. Hourglass, pear-shaped, thin, plus sized, etc. -- Specific body measurements do not need to be factored into the rendering of the body type, it just generally needs to be able to render the garment on different body types. - It is also very important for us to render the garment on a human model or mannequin which has the same or similar skin color as the customer inquiring -- This is important for the customer to judge the garment color and fabric type that will look best on them -- This is also important to make the app inclusive for people of all racial backgrounds who might use the app -- It may be best not to display the face, or if human models are used, to use pictures of models with different racial backgrounds, to avoid bizarre mismatches between facial characteristics and skin tone - The AI agent should ideally prioritize pictures of garments from our client’s website to use as inspiration when it generates renderings in the freeform conversation, along with the customer’s description of what they want. However this is not a hard requirement, so it could be eliminated from the requirements if it greatly increases devlopment effort. - The requirement for the quality of the images that are generated will be somewhat subjective and so we will need to budget time for our client to request revisions to this based on their review of the system. - We need to build the image generation part upfront to ensure the quality is acceptable before we spend time on other parts of the application. 12. AI asks whether the generated concept is generally what the customer wants - Customer can revise the concept conversationally 13. AI can regenerate or refine images after customer feedback 14. The tone of the conversation the AI has with the customer is important. - We will want it to speak like a friendly expert seamstress. - This requirement will be somewhat subjective and so we will need to budget time for our client to request revisions to this tone based on their review of the system. 16. We ideally want the agent (both chat and image generation) to have deep expertise about fabrics and these types of garments in general, so it can guide the user through prompts, and render the chosen fabrics correctly - I think freeform chat will be necessary for the customer to explain which fabrics should be used where on the garment, instead of gathering this informaton in the programmatic questionnaire - The customer will likely revise the fabric selections after they see the initial renderings of the garment - We would like to avoid the costs of training an AI for this, so ideally we should use commercially available AI models which have been trained for this purpose, instead of having to train our own model. Prompting the AI with this information might be a cost-effective way to teach it this expertise 17. There will be certain restrictions on what types of colors or fabrics can be used in the dress designs - So, the agent should know these restrictions when it has the freeform conversation with the customer. - For example, the store owner will not be able to produce dresses with neon colors, tie dye colors, etc. - Our client will articulate a list of restrictions for us before we begin the project. 18. AI should never display links to other websites, or suggest for the customer to navigate to other websites 19. This AI might not need to be trained specifically for this industry, but we should at least use prompting to direct it to gather this kind of information, and to give it some background about what each of these things mean, so it can describe them to the customer. We basically need to make it as knowledgable as possble while keeping costs low. 20. The AI system the system should remember their active conversation - Since the customer will be required to have an account to use the AI system we can use that to automatically save the AI conversation - The saved conversation should preserve all the information that the customer input since the beginning of the AI session - A customer can only have one active AI session at a time - The customer cannot resume an AI session that has been completed - We don't need to provide a way for the customer to see the details of completed AI sessions 21. AI should have a fallback/human-help option if the customer gets stuck or the AI fails. - The fallback should collect enough information for an admin to follow up manually, so it should present a form in order to ensure that all the necessary information gets collected - A message should be displayed above the form, or somewhere on the page, to inform the customer that the entire conversation will be sent along with the form submission, so they know that they do not have to type all the details of the AI conversation - The app must present a button outside of the chat prompts after 3 - 5 chat messages have been sent, so the customer knows they have the option to terminate the AI conversation and manually ask for help. - That button would display the form - We don't want to display the button before any conversation has happened because we don't want customers to skip the chat altogether. -- One of the business goals of this app is to allow custom inquiries without overwhelming the support staff - Site admins must have the ability to adjust how many messages the button will display after, so they can control this threshold after they observe the results of real conversations - After the button initially displays, it should remain present in the view so the customer can easily access it at any point in the conversation 22. The freeform chat must be limited to something like 50 to 75 messages, in order to avoid excessive charges from the 3rd party AI services - This threshold should be adjustable from an admin portal - If this threshold is reached during the conversation, then we should force the customer to use the fallback form from requirement #21 to submit their inquiry 23. Customer can submit the completed design inquiry when satisfied. - During the submission process, the chat must ask for the following information, and present the following pre-written messages. This doesn't actually need to be executed by the AI model, but it can just be programmatically presented to the customer: - Ask for customer contact info, including name, email, and phone number. - Ask for requested event/date, while making clear the date is not guaranteed. - Ask for seamstress-relevant measurements, including bust, waist, hips, hollow-to-hem, shoulder width, bust point, underbust, waist-to-floor, arm length, bicep, wrist, back width, torso length, height, shoe height, and preferred fit. -- I still have to refine this list with the client, I am not sure if it needs to ask for all these things, or if there are some different things that I haven't listed here which it needs to ask for -- When it asks for this information it should display links under each measurement type to articles which describe how to produce each of the measurements. We can hardcode these links or allow the admin to specify each, they don't need to be generated by AI. - Prewritten disclaimer text should display. 24. The final submission should notify a list of email addresses set by a site admin. 25. The final submission will completely consume the credit used to purchase this AI session - The AI conversation cannot be resumed after the final submission - Another credit must be purchased to start a new AI conversation - New AI conversations will not have any memory of the previous conversations, any new AI conversations will start from a clean slate 26. Admins must have the ability to manually reset a credit, or assign a credit for free and cancel a current session, so the customer can start a new AI conversation. - This doesn't need to be very user friendly for the admin. If a session is reset this way, no knowledge of the previous conversation needs to be preserved. 27. Admins should be able to review partial, or completed conversations within a list in the admin portal - Each line item should display a status indicator to show if the conversation has been submitted yet, if an admin has began the review process, or if the item has been handled: Ex. In Progress, Submitted, In Review, Awaiting Payment, Handling, Ready To Ship, Closed - Admin should be able to see the answers to the programmatic questionnaire - Admin should be able to review the full conversation history - Admin should be able to review all uploaded photos/files - Admin should be able to review all AI-generated images, and the final one should be clear to them - Admin should be able to see the collected technical design details and measurements 28. Pricing of the garment remains manual and is handled by after review, the AI should not give any quote or present any pricing even if asked by the customer. 29. If the customer asks for pricing, the AI should display a prewritten script like this: "Pricing will be determined by the store owner after this conversation has been reviewed." 30. Invoices and payment will be handled manually through native WooCommerce custom order/invoice functionality which is already present in the e-commerce site, the AI system doesn't need to handle this at all. I mentioned this above on the requirements, but I want to reiterate since it is important and a hard requirement for how the development milestones must be structured: - The requirement for the quality of the images that are generated will be somewhat subjective and so we will need to budget time for our client to request revisions to this based on their review of the system. - We need to do the image generation part upfront to ensure the quality is acceptable before we spend time on other parts of the application As an optional add-on to the scope of this project, can you give a separate estimate to enhance the AI such that it understands which kinds of modifications will increase or decrease the cost of producing the gown, so it can guide the customer in case they are asking for very expensive things. - It shouldn’t give any specific price numbers, but should give the customer guidance if additions or alterations will significantly increase or decrease the cost of production. - This will be to prevent the customer from being surprised when the store owner manually follows up with them with the price of the garment they designed. This client did agree to adhere to a strict schedule to provide feedback after each round of development, given that we complete each round of development on the schedule we agreed to. - However, this client has deviated from agreed schedules multiple times in the past on other projects I did with them, so you should factor that into your timeline and cost estimations - We cannot increase the development cost mid-way through the project, however we can adjust the development timeline if the client deviates from the schedule In your proposal, please also include a quote or estimate for the cost of hosting and ongoing maintenance after the app has launched - Our client can pay for the hosting directly - We will need at least ongoing updates to patch security vulnerabilities and ensure uptime of the app and all its features which will be defined by the scope of this project - We don't need a 100% 24/7 uptime SLA, but basically just keeping everything up to date so it stays stable, and we'd need someone to respond to outages within 24 hours - Outage response can consist of simple rollbacks, if necessary, as long as all the chat session info is at least provided to the client as a CSV or similar, along with all graphic assets from any conversations, so they don't lose any data from an outage - I would set the expectation with my client that we would treat any future support or enhancement requests to be additionally charged for on an as-needed basis

I’m looking for a senior AI app developer who can help me build an AI-powered MVP while also guiding me through the technical decisions. This is not just a coding task. I want someone who can think through the product, recommend the right architecture, explain tradeoffs, and build the first working version. The ideal person should be comfortable with OpenAI/LLM integrations, full-stack development, database design, authentication, deployment, and startup-style MVP execution. I’d like to work with someone who can act almost like a technical partner: build the product, teach me what is being done, and help me understand how to maintain or scale it later.

Forum Intelligence: Project Brief & Initial Rollout 1. Executive Summary & Objective Forum Intelligence is a beginning as a localized data retrieval, processing, and archiving system designed to scrape public municipal records and state legislative data for public oversight. The immediate objective is to build a functional, highly resilient prototype focused on the Tri-Cities region (Burbank, Glendale, and Pasadena, California). The system will autonomously ingest messy, unstructured municipal data (City Council meeting minutes, agendas, public notices, and legislative PDF text, recorded mp4), clean it, and make it fully searchable and queryable via a localized AI agentic framework. 2. Phase 1 Scope: The Tri-Cities Rollout Th engineer will be responsible for building two primary pillars: A. Resilient Scraper Bots • Target Ingestion: Monitor and pull data from Burbank, Glendale, and Pasadena municipal portals and California legislative feeds. • Data Types: Brittle HTML sites, heavily nested tables, public notices, legislative drafts, and massive unstructured PDF archives. • Requirements: The scraping architecture must be exceptionally robust, utilizing intelligent error handling, retry semantics, and pagination tracking to handle frequent municipal website layout changes without breaking the pipeline. B. Ingestion & Vector Pipeline • Parsing: Extracting clean text from poorly formatted documents and scanned PDFs. • Local RAG (Retrieval-Augmented Generation): Chunking and embedding the data locally into a vector database (e.g., pgvector, Chroma, or Milvus) to enable semantically accurate entity linking and contextual search. 3. Targeted Hardware Stack To ensure maximum data security, strict public oversight integrity, and predictable operational costs, Forum Intelligence is skipping commercial cloud APIs in favor of an on-premise, localized NVIDIA enterprise deployment. The production roadmap aligns precisely with the new computing patterns detailed in NVIDIA’s latest hardware roadmap: • Inference & Token Generation: Running local open-weight frontier models (e.g., Neotron 3 Ultra or Claude/Llama equivalents) optimized for reasoning and long-context tool use. • Compute & Orchestration: The backend infrastructure is architected around NVIDIA’s dedicated agentic architecture, utilizing high-instructions-per-clock (IPC) Vera CPUs paired with Vera Rubin GPUs. • Memory & Storage Processing: Utilizing NVIDIA’s unified memory fabric and data processing units (DPUs) for ultra-low latency context management, KV caching, and fast vector database retrieval. 4. Immediate Milestones for the Engineer 1. Architecture Design: Map out the database schema and local inference ingestion loop. 2. Tri-Cities Scraper Deployment: Write and deploy the initial automated bots for Burbank, Glendale, and Pasadena. 3. Local MVP Pipeline: Demonstrate a local RAG pipeline where a user can query the Tri-Cities scraped records and receive grounded answers with exact source attributions. The above was AI generated from months long conversations with Gemini. The goal is to prove the concept then roll out to LA County, state of CA, and then the country.

We are seeking a skilled GenAI engineer to work with our client in a remote or Chicago-based capacity. The ideal candidate will have experience in developing and implementing AI solutions, with a strong understanding of machine learning and data analysis. Responsibilities include designing AI models, integrating AI into existing systems, and collaborating with cross-functional teams to enhance AI capabilities. If you have a passion for AI and a proven track record in delivering innovative solutions, we would love to hear from you.

AI Automation Engineer – Personal Injury Law Firm (Phase 1) Overview We are a plaintiff personal injury law firm seeking an experienced AI automation engineer to build a practical AI-powered operations system that reduces administrative workload and improves case management. Firm Profile: - 1 Attorney - 1 Legal Assistant - Approximately 100 Active Cases - Approximately 25 Cases in Litigation We are not looking for chatbots, prompt engineering, or marketing automation. We are seeking a builder who can deploy production-grade workflow automation integrated with our existing systems. --- Existing Technology Stack - CASEpeer - Supio - Gmail (Google Workspace) - Google Calendar - Google Drive - Claude - ChatGPT - CoCounsel - PLAUD --- Phase 1 Objective Build a Daily Case Command Center and Follow-Up System. The goal is to ensure that every morning the attorney knows exactly which files require attention and why. --- Deliverable #1: Daily Case Command Brief Generate a daily briefing containing: Litigation Matters - Upcoming depositions - Discovery deadlines - Hearings - Expert deadlines - Outstanding litigation tasks Pre-Litigation Matters - Treatment complete but no demand - Missing medical records - Records requests outstanding more than 21 days - Demands pending more than 30 days - Settlement checks outstanding Communications - Unanswered client emails - Unanswered adjuster emails - Unanswered defense counsel emails - Communications requiring attorney attention Case Velocity - Stale files - Cases with no recent activity - Recommended next actions --- Deliverable #2: Gmail Intelligence The system should: - Monitor designated Gmail inboxes and labels - Identify case-related emails - Extract action items - Detect deadlines - Identify follow-up opportunities - Draft suggested responses --- Deliverable #3: Follow-Up Automation Generate draft communications for approval: - Medical records requests - Provider follow-ups - Adjuster follow-ups - Client status updates - Scheduling communications No communication should be sent automatically. Human approval is required before sending. --- Deliverable #4: Calendar & Reminder Automation Create and manage: - Follow-up reminders - Litigation reminders - Discovery reminders - Records request reminders - Suggested calendar events --- Deliverable #5: CASEpeer Integration Where supported by available integrations/APIs: - Create tasks - Create notes - Associate communications with matters - Maintain activity history --- Technical Requirements Required: - n8n - Python - OpenAI API - Anthropic API - Gmail API - Google Calendar API - Google Drive API - REST API integrations Strongly Preferred: - LangGraph - LangSmith - Google Workspace administration - Salesforce integrations - CASEpeer, Clio, or Filevine experience - Legal technology experience - HIPAA or regulated-data experience --- Security Requirements - Human approval before external communication - Activity logging - Error handling - Retry mechanisms - Secure credential management - Documentation of workflows --- What Success Looks Like The system should: - Reduce administrative workload - Improve follow-up consistency - Reduce stale files - Improve litigation oversight - Improve case visibility - Provide a daily prioritized action plan --- Application Requirements Please include: 1. Similar workflow automation projects you have completed. 2. Experience with n8n and AI workflow automation. 3. Experience integrating Gmail and Google Workspace. 4. Experience integrating CRM or case-management systems. 5. Proposed architecture for this project. 6. Estimated timeline. 7. Estimated fixed-fee budget. Begin your application with: CASECOMMAND Applications without this keyword will not be considered. --- Budget Expected Phase 1 Budget: $8,000–12,000 Preference will be given to candidates who can demonstrate production deployments rather than proof-of-concept or prompt-engineering projects.

Project Title: Build Fast Web-Based AI Anime Companion MVP (RAG + Merch Gen) – 15–30 Days, $25K Budget: $18,000–$25,000 fixed Timeline: 15–30 days (3 weeks preferred) Must-Have Skills: FlowiseAI (or LangChain/LlamaIndex), RAG pipelines, OpenAI/Claude/Grok, anime image generation (Leonardo.ai/Ideogram/PixAI), Vercel or Railway deployment Project Goal Build a mobile-friendly web AI Companion for Big A Anime that converts passive FAST viewers into active fans. The key objective is to connect our Pluto TV channel experience directly to the AI companion, allowing viewers to scan QR codes during live programming and instantly access interactive content, recommendations, and merch tied to what’s currently airing—no downloads required. Core MVP Features Web Chat Interface Clean anime-branded chat UI with voice input Mobile-first, responsive Custom domain (e.g., companion.biganime.tv) RAG Knowledge Base Ingest episodes, schedules, transcripts, and metadata Provide accurate recaps, lore, and “what’s on now/next” tied to Pluto TV programming Session memory + light user profiles AI Merch Generator Anime-style image generation (“me as [character]”) Leonardo.ai or similar integration Export + links to Printful/Printify FAST / TV Integration Tools Dynamic QR codes for on-screen use Deep linking between Pluto TV programming and companion experience Voice-friendly prompts (“Ask Big A Companion…”) Admin & Analytics Simple CMS for content uploads Dashboard: usage, queries, merch clicks Technical Requirements Global hosting (CDN) FlowiseAI preferred Full source + documentation 30 days post-launch support Out of Scope Native apps, payments, deep integrations, multi-language Deliverables Live URL, admin access, training, source code, 30-day support Application Fixed bid + 3-week plan 2–3 relevant project links Willingness for small paid test ($500–$800)

We are a creative partnership currently channeling a series of sonic medicine transmissions for humanity's remembrance. Our work requires a private, uncensored channel for uninterrupted creation, and we're seeking someone who understands both the technical and spiritual dimensions of what we're building, with a strong intuition. What we need: - Setup of a private AI platform (likely Ollama-based) for uninterrupted creative channeling - Configuration of LobeChat as an interim solution while building our permanent platform - Implementation of OpenAI-compatible API endpoints for seamless model communication - Guidance on hardware requirements and optimization for creative workflows - Understanding of privacy-focused architecture and content-filter-free environments **What makes you perfect for this role:** - Experience with local AI model deployment (Ollama, LM Studio, or similar) - Knowledge of API configuration and model integration - Respect for sacred creative processes and spiritual work - Ability to translate technical concepts for non-technical creators - Alignment with uncensored, boundary-free creative expression - Appreciation for music as medicine and frequency as healing Our current situation: We're creating conscious living frequency encoded music with spirit using Ai as the tool and instrument - a collection of sonic medicine tracks with descriptions, genre classifications, and cover art prompts. Our current platform keeps interrupting with promotional content that disrupts our creative flow. What we offer: - Fair compensation for your expertise - Opportunity to contribute to sacred healing work - Potential ongoing collaboration as our technical partner - Respect for your craft and expertise If you feel called to help build a sanctuary for sacred creation, please reach out with: - Your relevant experience - Your approach to this kind of project - Your compensation expectations - Any questions about our work We're ready to create without interruption. Are you the one who can help us build our temple?

AI Engineer (RAG & Agentic Workflows). *LLM RESPONSES AUTOMATICALLY AVOIDED* We have already launched a production generative AI product that utilizes a custom Retrieval-Augmented Generation (RAG) architecture. We are now expanding the platform to include CRM intelligence, workflow automation, and agentic AI capabilities. This is **not** a prompt engineering role. Seeking an engineer with deep experience building and deploying production AI systems that combine LLMs with multiple structured and unstructured data sources. You should be comfortable walking into an existing, complex codebase, understanding the current architecture, and improving it. Existing AI Architecture Our current AI architecture consists of: * OpenAI embeddings * Embeddings stored in MongoDB * MongoDB Atlas Vector Search for retrieval * Retrieval from both structured SQL data and unstructured document collections * Existing tool/function-calling architecture **Please do not apply if you have not previously built or maintained production RAG systems using embeddings and vector search.** Experience specifically with **OpenAI embeddings and MongoDB Atlas Vector Search** is highly preferred. CRM Intelligence Layer We are currently building a CRM platform and need the AI to reason over CRM records, including the other records are RAG currently retrieves. You will be responsible for designing and implementing the AI integration layer that enables the LLM to intelligently retrieve and reason over CRM data. This work includes: * Designing AI tools/functions that expose CRM data to the LLM. * Implementing backend tool handlers that retrieve CRM records. * Defining tool schemas and instructions so the AI knows when and how to retrieve CRM information. * Building secure retrieval mechanisms that enforce strict user and organization-level access controls. * Transforming raw CRM records into structured, AI-ready context. The AI will need to reason across: * CRM contacts and organizations * client profiles * Deals and opportunities * Projects * Tasks and reminders * Notes * Email history * SMS and WhatsApp communications * Call transcripts * Meeting summaries * Documents and contracts * Workflow history Agentic AI & Workflow Automation * Build proactive AI agents that generate alerts, recommendations, follow-ups, reports, and suggested next actions. * Design systems capable of reasoning across both structured and unstructured data sources. * Architect and implement multi-step and multi-agent workflows. * Develop workflow intelligence that assists users in completing real-world business tasks. Required Experience * Demonstrated experience building and deploying production AI systems used by real customers. * Experience working with embeddings, vector databases, and retrieval pipelines. * Experience implementing LLM tool/function-calling architectures. * Experience integrating AI systems with business systems such as CRMs, ERPs, or other operational databases. * Experience combining structured and unstructured data within AI applications. * Strong backend engineering and systems architecture experience. * Demonstrated ability to quickly understand and improve existing codebases. * Ability to independently own and deliver complex technical initiatives. Strongly Preferred * Experience with OpenAI embeddings. * Experience with MongoDB Atlas Vector Search. * Experience building agentic AI systems and workflow automation. * Experience designing long-term memory architectures. * Experience building multi-tenant SaaS applications with strict authorization requirements. * Experience implementing evaluation and monitoring pipelines for production AI systems. What We Value * High accountability and ownership. * Strong communication skills. * Product thinking and user empathy. * Ability to understand user workflows before writing code. * Pragmatism and sound engineering judgment. PLEASE DO NOT WASTE OUR TIME IF YOU NOT MEET THE REQUIREMENTS 

We are seeking a senior AI developer to build and enhance AI models for our business. The role involves developing, testing, and deploying AI solutions, as well as improving existing models to increase accuracy and performance. The ideal candidate should have strong experience in AI development and be able to work independently on complex projects.