NOBS Marketplace: AI that finds, scores and contacts link prospects
A marketplace where an AI agent does the prospecting and outreach, and people approve what matters.
Platform
NOBS Marketplace · AI SaaS
Engineering
Led by our CTO
Website
nobsmarketplace.com
Type
AI SaaS · Agentic workflow
Overview
Link building is slow, repetitive work: finding relevant sites, judging their authority, writing personalised outreach and running multi-step campaigns.
NOBS Marketplace hands that work to an AI agent. The Smart Campaign agent orchestrates every stage, while people stay in control of the decisions that matter. Lunova's CTO led the engineering of the platform.
The Problem
Useful AI needs to be safe, not just impressive.
Enterprises want AI that does real work inside real systems, without running up costs or sending something nobody approved.
That takes more than a prompt: grounded knowledge, connections to CRM, search, mail and analytics, and controls that keep an agent inside its limits.
The Solution
An agent wired into real systems, with guardrails from day one.
The Smart Campaign AI agent uses RAG, a vector database and tool calling. MCP servers connect it to CRM, search, mail and analytics, so it acts on real data instead of guessing.
AI guardrails cover evaluation, human approval, and cost and rate controls. Payments run on Stripe Elements with saved cards, and cards can be shared across an organisation.
How It Works
Five stages, orchestrated by one agent.
Prospect discovery
The agent finds candidate sites that could link to a customer.
Relevance & authority scoring
Each prospect is scored so effort goes to the best opportunities.
Personalised outreach
Outreach is written for each prospect rather than sent as a template.
Multi-step campaigns
Follow-ups run as structured, multi-step campaigns.
Human approval
People approve the key decisions before anything goes out.
What It Proves
We ship AI agents with guardrails, not demos.
NOBS Marketplace runs agentic AI in production, connected through MCP, with evaluation, approval and cost controls built in from the start.
It is the pattern we bring to client AI work: a clear business target, an agent that can act, and limits that keep it safe.
"Enterprises want AI that is useful and safe. That means guardrails built in from the start."
- Lunova Digital
Have a similar problem?
We build software that solves real operational problems.