Haystack RAG Pipeline
Budget: €250 – €750 EUR
Developer Wanted – Deploy Haystack RAG Pipeline for Modular AI SaaS (Childcare Sector)
We’re building a modular AI SaaS platform for the childcare industry. Our backend is built on a Haystack RAG architecture, and we’re looking for a developer who can take our MVP setup to full production — including secure multi-tenant support and multiple functional AI modules.
________________________________________
Your responsibilities:
• Deploy and productionize our existing Haystack RAG pipeline
• Integrate the backend with our frontend stack (Vue + Electron) via API (FastAPI or Node)
• Implement and modularize the following AI features:
o Document Chat — includes source highlighting and references in AI responses
o compliance check
o Patron analyses
o Document analysis
• Implement role-based document filtering using Keycloak JWT tokens (per tenant/role)
• Set up logging, error handling, and fallbacks for missing data
• Ensure the system is scalable and maintainable for future growth
________________________________________
Bonus if you have experience with:
• Extracting and analyzing data across multiple documents to produce coherent summaries or analytical responses
• Answering questions with source-linked citations and highlighted reference text
• Dynamic prompting, chaining logic, or advanced query routing within Haystack
• Working with Vue, Electron, and Keycloak in multi-module applications
________________________________________
Our tech stack:
• Haystack 2.x (RAG pipeline)
• FastAPI (Python)
• Vector DB (e.g. Elasticsearch)
• Open-source LLMs or GPT (via API)
• Keycloak (multi-tenant SSO, JWT auth)
• Vue 3 + Electron (frontend)
• Docker (optional, for deployment)
________________________________________
You are:
• Experienced with Haystack or similar RAG/retrieval frameworks
• Skilled in modern Python development (FastAPI, Pydantic, async)
• Comfortable working with multi-tenant authorization models
• Able to structure modular code that supports future expansion
• Someone who values clean, documented code and a thoughtful product vision
Collaboration and practical feedback appreciated
Please include relevant examples of past work or a short motivation. We're looking for someone who understands both technical architecture and real-world usability.
We’re building a modular AI SaaS platform for the childcare industry. Our backend is built on a Haystack RAG architecture, and we’re looking for a developer who can take our MVP setup to full production — including secure multi-tenant support and multiple functional AI modules.
________________________________________
Your responsibilities:
• Deploy and productionize our existing Haystack RAG pipeline
• Integrate the backend with our frontend stack (Vue + Electron) via API (FastAPI or Node)
• Implement and modularize the following AI features:
o Document Chat — includes source highlighting and references in AI responses
o compliance check
o Patron analyses
o Document analysis
• Implement role-based document filtering using Keycloak JWT tokens (per tenant/role)
• Set up logging, error handling, and fallbacks for missing data
• Ensure the system is scalable and maintainable for future growth
________________________________________
Bonus if you have experience with:
• Extracting and analyzing data across multiple documents to produce coherent summaries or analytical responses
• Answering questions with source-linked citations and highlighted reference text
• Dynamic prompting, chaining logic, or advanced query routing within Haystack
• Working with Vue, Electron, and Keycloak in multi-module applications
________________________________________
Our tech stack:
• Haystack 2.x (RAG pipeline)
• FastAPI (Python)
• Vector DB (e.g. Elasticsearch)
• Open-source LLMs or GPT (via API)
• Keycloak (multi-tenant SSO, JWT auth)
• Vue 3 + Electron (frontend)
• Docker (optional, for deployment)
________________________________________
You are:
• Experienced with Haystack or similar RAG/retrieval frameworks
• Skilled in modern Python development (FastAPI, Pydantic, async)
• Comfortable working with multi-tenant authorization models
• Able to structure modular code that supports future expansion
• Someone who values clean, documented code and a thoughtful product vision
Collaboration and practical feedback appreciated
Please include relevant examples of past work or a short motivation. We're looking for someone who understands both technical architecture and real-world usability.