We connect n8n, Make or Zapier with AI models: cart recovery, chatbot, stock alerts and customer service. We hand over the configuration of the implemented flows.
Four things always included in your AI automation package from Balance IT.
Current process brief + automation list → quote within 24h.
Export JSON / YAML at any time. No vendor lock-in. Self-hosted (Docker) option.
OpenAI / Claude / Gemini / open-source (Llama) models. You choose in the brief, we implement.
Every run is logged — when, what it did, what it cost (LLM tokens).
Email flow, chatbot, inventory alerts, lead scoring, customer service — everything runs 24/7 without your involvement.
Audit of current processes (automation list + priorities)
Workflow design + AI prompts (your persona / branding)
Build + testing + integrations with an existing shop / CRM
Workflows with exportable JSON — no vendor lock-in.
Choice per use case. Token + cost logs.
Docker option — your data on your server.
Every run is logged. Slack / email alerts on errors.
Choose the pace and scope — always at a fixed price, with exportable workflows. Net prices. The stated amounts cover the package scope; additional work requires an agreed scope and a confirmed quote.
Implementation: 1-2 weeks
Implementation: 3-4 weeks
Custom quote
Proven tools, exportable workflows, no vendor lock-in — take out any component at any time and host it yourself.
Chatbot with RAG (your offering, FAQs, documentation). Shop widget integration + Slack + email.
Abandoned cart, win-back, post-purchase sequence. LLM personalisation (welcome, recommendation).
Trigger when stock < threshold → supplier email + team Slack + draft PO. AI checks history for the optimal quantity.
Orchestration platform — with exportable JSON workflows. Self-hosted (Docker) option for privacy.
Every form lead is scored by an LLM (intent + budget + timeline). CRM tag + Slack alert for hot leads.
Every run is logged (input, output, LLM token cost). Error alerts. Monthly cost vs value report.
We review these questions with you before we confirm the scope and deliverables.
Scope and handover
How should catalogue data, prices and stock levels stay up to date?
How should payment, ordering and returns work?
What data should go to payment, stock or ERP systems?
How will we test the first order and hand over store operations?
The Balance IT standard
Export JSON / YAML at any time, no vendor lock-in.
Quote within 24h, invoice only after you approve the stage.
Every run logs LLM cost — monthly cost vs business value report.
Workflows + DB + LLM proxy on your Docker setup. Full data control, no vendor lock-in.
Every stage has its own team and success metrics. This makes AI automation projects as carefully developed as the homepage.
Each stage ends with a demo and approval. Only then do we invoice and move on.
Week 1 · workshop + map
Week 2 · flowchart, prompts, mock-up
Week 3 · n8n / Make + integrations
Week 4+ · ongoing
Loading projects...
Answers to frequently asked questions about AI in E-commerce