Enterpise level project help

Hi all, I’m in a bit of a pickle. A client reached out to me today about building what is essentially a SaaS product: a custom web app with user authentication, project management-style functionality (workflows, timelines, multiple modules), and an integrated LLM.

Security is a high priority (I’m an IT/network guy, and my co-founder is a SWE). They specifically do not want a “ChatGPT wrapper.” They want a private LLM deployment with high assurances that company data and secrets aren’t leaked, which likely means a self-hosted or highly controlled deployment. They also want us to manage and maintain the system every month.

My questions are:
Is it even possible to build in-house using n8n as part of the tech stack, or is this a fully custom build-out?

Is it possible? What would you realistically charge for something like this (initial build and monthly management)?

P.S He wants to sign me up under an NDA, then discuss further, as this is all he’s told me so far

And I’m happy to try, but want to know if this is outside our skillset

Hi @zamonte

  • Yes, it is possible to use n8n as part of the stack
  • No, n8n cannot be the core of the SaaS application
  • This would be a custom SaaS build with:
    • A real backend + frontend
    • A controlled / private LLM deployment
    • n8n used behind the scenes for orchestration, automation, and integrations
  • This is well within reach for a strong IT + SWE team if scope is controlled
  • Pricing is mid–high five figures minimum for build, and four figures/month+ for ops

If you try to build this as an “n8n app”, it will fail.
If you treat n8n as infrastructure glue, it can be a very strong choice.

What actually works in production looks like this:

[ Web / Mobile App ]
        |
        v
[ Custom Backend API ]
(Auth, RBAC, business logic)
        |
        v
[ n8n (self-hosted) ]
- LLM calls
- Long-running workflows
- Data enrichment
- Background jobs
- Audit logging
- Integrations
        |
        v
[ LLM Stack (private) ]
- Self-hosted (e.g. vLLM / Ollama / TGI)
- Or controlled provider (Azure OpenAI w/ private networking)
1 Like

Hi @zamonte , For the question Is it even possible to build in-house using n8n as part of the tech stack, or is this a fully custom build-out? , Yes i can say that you can build that kind of ‘chatgpt’ wrapper using n8n as a back-end infrastructure and it is going to save you a lot of time also and the production quality would be commendable.

And as you have asked Is it possible? What would you realistically charge for something like this (initial build and monthly management)? It depends and let me tell you how it depends , as these projects are not small nor their complexities so you first consider your cost per/hour and how much time it is going to take to build completely, now i would say instead of charging the client as a 30% first 70% later deal you should charge them on retainer and also it is going to depend they are giving you the infrastructure i mean N8N cost, AI Models, Data bases. So divide your project in months and now this is going to make the client feel ease that they do not have to pay you a 4 digit check, and alongside this you should say that we are going to keep a tracker and a plan that client can see that what you are going to build and how far you have came with estimated dates and layout.

Before making this retainer statement consider calculating the ROI of the product you are building for them with months saved in time and cost, like a report before contract to showcase that this is what they are gonna get if they pay you. If you are looking for a static amount for your monthly retainer please do calculate it with your PROFIT margin and hourly rate.

good job! Now it is your time to build a small plan just a high level overview of things are gonna work so that you can showcase that to anyone not just an IT guy, let me know if you need assistance with that. happy to help!

1 Like

@zamonte I can tell you how we handle this. In the case of firms that have data privacy as a major requirement, we have custom llm on their own private vps. A new vps for their firm, services like n8n, local llm, supabase locally.

Such projects are interesting to work on. Usually high paying. They require some research before giving a final proposal to the clients.

2 Likes

Hi Zamonte,

Yes, we can use n8n for automating workflows and integrations, but the core app with user authentication and project management features would require custom development (likely with Node.js, Python, or similar). For LLM deployment, it would need to be self-hosted or controlled, with n8n handling interactions.

if your team has custom web dev and security expertise, but for LLM deployment, consider partnering with a specialist.

Let’s discuss more after the NDA.

1 Like

This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.