Where Kommo's marketplace ends, we build.

Often a widget covers 70–80% of mid-size SMB needs. Custom is for the 20–30% where standard widgets stop short: ERP sync, custom AI workflows, in-card calculators, multi-system orchestration. We tell you honestly during scoping which path fits your case.

The 30-minute scoping call is free and produces a rough cost band on the call. The technical discovery that follows is paid (fee disclosed on the scoping call) and produces a written specification with architecture diagram, data-flow map, risk register, and a fixed-fee project quote. You can accept it, reject it, or take it to another integrator — no lock-in.

Yes. Source code is delivered to your repository on completion. We retain rights to reuse generic patterns (e.g., a Stripe-payment-link template), but the integration code specific to your business stays with you.

We pause and re-scope. We do not silently expand work and bill it. About one in three projects gets a re-scope mid-build, almost always because new edge cases surface during sandbox testing. The conversation is: here's what we found, here's what it changes, here's the new estimate, you decide.

Yes. Many engagements are co-built with the client's own engineering team. We can lead, co-lead, or just architect-and-review while your team writes the code. Clean handoff boundaries by default.

On a maintenance retainer we monitor Kommo platform updates and patch your integration before users notice. Without a retainer we offer reactive bug-fixes at standard hourly rates after the 30-day post-delivery hyper-care period ends. Retainer tiers and fees set per engagement.

Every integration ships with a signed DPA from WeAreWay Ltd. UK-GDPR and PII handling are baked in by default; HIPAA-aligned controls available on request for healthcare engagements. For chat-channel AI use cases — see our /services/ai page (Kommo's native AI Sales Agent). For AI workflows that go beyond Kommo's native agent — e.g. custom embeddings, RAG, voice transcription pipelines — we use enterprise APIs from OpenAI or Anthropic in EU regions, with training opt-out and zero-retention settings where the provider supports them. Choice of model is documented in your DPA.

Typical sequence: 1 week scoping → 2–12 weeks build → 1 week UAT and deployment → 30 days hyper-care. Small integrations finish in 1–2 weeks total. Large multi-system orchestration runs 6–12+ weeks. We never start construction without a signed scope and fixed-fee quote.