Häufig gestellte Fragen
Konkrete Antworten.
Was Kunden fragen, bevor sie sich melden, mit den Antworten, die wir in einem ersten Gespräch geben würden.
Diese Seite ist derzeit nur auf Englisch verfügbar. Übersetzungen folgen.
- What does Sophonix do?
- Sophonix B.V. builds AI-driven business automation for SMB and mid-market companies. The work has three legs: business analysis (deciding what's worth automating before anyone writes code), workflow design (end-to-end, including the boring edge cases automations break on), and AI implementation (build, run, and stay around long enough to make sure it sticks).
- Who do you work with?
- Small-to-mid-sized businesses, typically 10–150 employees with EUR 1–25M in revenue, based in the Netherlands or broader EU. Buyers are usually owners, managing directors, or operations leads — people who are ChatGPT-curious but not technical themselves. Sophonix does not target enterprise rollouts, VC-backed startups, or government contracts.
- What industries do you serve?
- Industry-agnostic. Sophonix has built across martech (Riviants), agentic browsers (Obscura), and algorithmic trading (Kaizen X and Kaizen FX). Client work follows where automation pays back — operations-heavy businesses with repetitive workflows that have real edge cases.
- How does an engagement work?
- Three phases, mapped to the three legs. Phase one is analysis: we look at your business and identify what's worth automating before anyone writes a line of code. Phase two is design: we map the workflow end-to-end including the boring edge cases. Phase three is implementation: we build it, run it, and stay around long enough to make sure it sticks.
- What does it cost?
- Scope drives the number. Sophonix works in defined-scope phases rather than open-ended retainers — you know what you're paying for and what you're getting before each phase starts. Email [email protected] with a rough description and you'll get a real estimate back.
- What's your tech stack?
- Whatever fits the problem. Often LLM-based agentic systems built on Claude or GPT-4o, structured data layers (Neo4j for graph relationships, Qdrant for vector recall), Python and TypeScript orchestration, deployed on infrastructure the client controls. Bias toward bare-metal or self-hosted where the data sensitivity warrants it — the same principle that drives Sophonix's own internal tooling.
- How long do projects take?
- Depends on scope. A focused business analysis sprint can be one to two weeks. A full implementation typically lands inside 8–12 weeks for a defined workflow. Sophonix does not chase 6-month enterprise rollouts — if a project can't ship inside a quarter, the scope is wrong and we re-cut it.
- Where are you based?
- Amstelveen, Netherlands (Laan van Kronenburg 14, 1183 AS). Sophonix works remote-first across NL and EU, async with US-based collaborators where engagements call for it. Site visits when the engagement benefits from in-person time.
- What makes Sophonix different from other AI consultancies?
- Sophonix builds for clients the way it builds for itself. The internal tooling that runs Sophonix's own work — BenBrain, a persistent memory layer running on bare-metal EU infrastructure behind Tailscale — is the same kind of durability work we'd do inside a client engagement. We don't ship demos; we ship systems we'd trust on our own deals.
- How do I get started?
- Email [email protected]. Tell us what you do, what's painful right now, and what a good outcome looks like. We come back with whether we think we can help — and if we can't, who probably can.
Frage, die wir nicht behandelt haben? E-Mail an [email protected].