Synthflow Alternative UK: Region Move Guide 2026

A UK team running an enterprise Synthflow workspace on the legacy Global cluster has a decision on its desk this quarter, whether or not it asked for one. Synthflow is retiring its legacy Global cluster, and its own migration notes say that by the end of 2026 every enterprise workspace still on that cluster moves to a dedicated US or EU cluster. An account that does not answer the region email lands in the US. For a contact centre in Leeds or a claims desk in Bristol, that is the moment to ask a wider question: stay, move region, or move platform.

This guide is written for the Head of Customer Operations who owns those calls, with the CIO as co-reader. It covers the cluster change, where a UK team's numbers really live on Synthflow, what an agent export carries and what it leaves behind, and how to move without dropping calls. It deliberately cedes three neighbouring topics to other pages. Pricing and certifications belong to our head-to-head comparison page, the five-vendor matrix belongs to the best AI voice agent 2026 guide, and the PECR and Ofcom duty teaching belongs to the Vapi alternative guide. The wider context is familiar: McKinsey's State of AI 2025 found 88% of organisations use AI in at least one function, but only 33% have it in production. A voice agent that already takes live calls is a production system, which is why a vendor-set platform change deserves a plan rather than a reply-all.

This guide is shipped by the team behind Dilr Voice, a multi-agent voice AI platform that chains specialised agents into a single phone call. Or see DATS, our five-stage method for placing AI where the operating numbers move.

What is changing on Synthflow before the end of 2026?

Synthflow is retiring its legacy Global cluster during the fourth quarter of 2026. Every enterprise workspace still on it moves to a dedicated US or EU cluster, chosen through a region email Synthflow sent through October. If an account does not reply, Synthflow moves the workspace to the US cluster. Changing region after the move is a manual, chargeable migration, and Synthflow warns that not everything carries across that way.

Synthflow's data region documentation explains why the change matters. The platform runs three clusters. Global stores data in the EU but can process live calls in the US or the EU depending on traffic and routing. The US cluster keeps storage and compute in the United States, and the EU cluster keeps both in the European Union. The same page is blunt about choice: the cluster picked when a workspace is created is permanent, and the only exception is the Synthflow-run retirement itself. Retell users faced a different vendor-set deadline this year, the retirement of multi-prompt agents, which our Retell AI alternative guide works through.

The migration notes split the work by destination. A move to the US cluster is light for the customer: Synthflow migrates agents, call logs and actions, the API base URL keeps working, and the only network task is adding the US egress and SIP addresses to any allowlist. A move to the EU cluster asks more. The customer repoints its API base URL at an agreed date and time, accepts a short coordinated switchover window, and, if it connects its own Twilio account, supplies a Twilio auth token issued in Twilio's Ireland region. The docs set Friday 9 October 2026 as the date for that token, which has now passed. In our view, an EU-bound team that missed it should raise it with its Synthflow contact this week rather than wait for the switchover slot.

Two scoping points keep this honest. First, Synthflow's migration page applies to enterprise customers on the legacy Global cluster; a workspace already created in the US or EU cluster is described as permanent in its region, and the page does not address non-enterprise workspaces. Second, nothing here says the US or EU cluster is a bad home. It says the choice is now explicit, has a default, and is expensive to reverse. That is the trigger for a structured review, the same review our AI placement diagnostic runs before any platform decision.

Why does the default region matter for a UK team?

The default matters because a UK business choosing where its call data sits cannot leave that choice to a vendor default. Under UK GDPR, whoever initiates a restricted transfer must cover it with adequacy regulations, appropriate safeguards or an exception, and the region decides where recordings and transcripts end up. A silent default to the US does not remove those questions; it means the region was chosen without a recorded decision.

The ICO's brief guide to international transfers, last updated on 15 January 2026, puts the responsibility plainly: "If you initiate a restricted transfer, you’re responsible for complying with the transfer rules." The ICO adds that this applies wherever an organisation sits in the processing chain, so the vendor and its sub-processors can carry transfer duties of their own. In our reading, the UK business that picks the platform and the region still owns the decision it can control, which is why the region email belongs with the data protection lead and not only with IT. The same guide sets out the three routes that can cover a restricted transfer: UK adequacy regulations, appropriate safeguards, or an exception.

An EU cluster does not make the question disappear either: data leaving the UK is still leaving the UK, whichever cluster receives it. Ask Synthflow, in writing, which entity receives the data, which transfer mechanism covers each region, and which sub-processors sit behind it. We do not repeat the transfer mechanics here: our international transfers guide covers the IDTA and transfer risk assessments, and our data residency guide covers hosting choices for enterprise voice.

What this section adds is the operational point. Synthflow's Twilio integration notes show that even an EU workspace does not keep every request in Europe. With a customer's own Twilio account connected, calls, recordings, SIP trunking and number lookup run in Twilio's Ireland region, but post-call SMS, in-call SMS, WhatsApp messaging and number purchase fall back to Twilio's US region. A UK team that sends appointment confirmations by text after each call should map that flow before it writes "EU hosted" in its records. The AI voice compliance guide for the UK and EU is the right companion for that mapping exercise.

Where do your UK numbers actually live on Synthflow?

On Synthflow, a UK team's production numbers are likely to live outside Synthflow already. Synthflow sells numbers directly only in the US, Canada and Australia, and its own guidance recommends Twilio or SIP for production so customers keep control of numbers and regional rules. A UK number is therefore usually held in the customer's Twilio account or with its carrier, which can turn a platform move into a routing change rather than a port.

Synthflow's telephony overview lays out the three ways in: buy a number in Synthflow, bring a Twilio account, or connect a carrier or PBX over SIP. The phone numbers page confirms the purchased range is limited to US, Canadian and Australian numbers for quick testing. Two consequences follow for a UK buyer. The SIP and PBX route needs a Synthflow Enterprise plan, so a team connecting a UK carrier is already on enterprise terms. And because the numbers are not Synthflow's, moving them is a matter of repointing the Twilio voice configuration or the SIP trunk at a new platform.

This is also where the vendor's positioning needs a careful reading. Synthflow's homepage describes it as the only "end-to-end Voice AI platform with in-house telephony", and its telephony docs say it runs its own session border controllers and media infrastructure. That is a real engineering choice about how calls are carried. It is not the same as owning a UK number range, and for a UK team the numbers callers know are usually held in the customer's own account. Ask your Synthflow contact to confirm, in writing, which numbers on your workspace are Synthflow-purchased and which are imported. If you would rather read the workspace with us, book a short scoping call.

The calling-line duties that apply on any platform are covered in the Vapi alternative guide linked above, and the outbound consent rules in our PECR guide to AI outbound calling. On Dilr Voice, telephony runs on Twilio or Vobiz, and outbound campaigns require the customer's own Twilio or Vobiz account. See the Dilr Voice product page for the full telephony note.

What does a Synthflow agent export carry, and what does it leave behind?

A Synthflow agent export carries the agent's configuration but not everything the agent does. Synthflow's version control can export any saved version to a JSON file, yet its documentation states that exports do not include actions or knowledge bases, and that the REST API does not expose version history. A team leaving Synthflow should therefore treat actions and knowledge bases as work to rebuild, not files to copy.

Synthflow's version control documentation is useful precisely because it is specific. Every publish saves a version, any version can be restored or compared, and Export to JSON downloads a version's configuration for import elsewhere through JSON import. Agents built in the visual Flow Designer, with nodes and branches, or in Prompt Builder, with one main prompt, both export this way. The gaps are the action layer and the knowledge layer: actions, the steps that transfer calls, book meetings and call your APIs, sit outside version history, and Synthflow tells users to re-add actions and knowledge bases after importing.

The data layer has its own rules. Synthflow's security and compliance settings offer a toggle that deletes transcripts, recordings and caller IDs after 30 days, and PII redaction that applies to new calls only, with existing transcripts and logs not redacted retroactively and live audio unaffected. Its call data export offers a manual CSV download from an agent's call history, while automatic daily exports to Amazon S3, Google BigQuery or SFTP are an Enterprise plan feature. Before any move, decide which call history you are legally required to keep, export it, and confirm the retention toggle will not delete it first; the storage questions in our data residency guide apply to that archive too.

The test assets deserve a sentence of their own. Synthflow's BELL loop (Build, Evaluate, Launch, Learn) puts real weight on its Test Center, where simulated conversations are scored against success criteria. Those suites are valuable and they are platform-shaped. Rewriting the scenarios as plain test cases you own, a habit our evaluation harness engineering notes argue for on any platform, is what lets a team compare old and new behaviour fairly during a move. Treat any asset not on this list as one to inventory by hand before cutover.

Should a UK team stay on Synthflow or switch?

A UK team should stay on Synthflow when its agents depend on the visual Flow Designer, it resells agents under its own brand, and the region decision lands cleanly in the EU cluster. It should consider switching when neither cluster fits its data requirements, when its agents depend on actions that need rework anyway, or when the enterprise commitment no longer matches call volume.

Staying has a real case, and our own head-to-head comparison page says so: Synthflow's drag-and-drop flow builder is the strongest no-code experience in this category, and reselling voice agents under an agency's own brand is a motion Synthflow supports and DILR does not. A team whose operations staff, not engineers, maintain the call flows should weigh that heavily. Synthflow says it migrates agents, prompts, knowledge bases and call history for enterprise workspaces, so a region move inside the same vendor asks far less of the customer than a platform change, which rebuilds actions and knowledge bases.

Switching has a real case too, and it rests on the facts above rather than on features. If a UK team is already reopening its transfer records and region choice, and its agents need reworked actions for other reasons, it is doing much of the groundwork of a platform evaluation. That is the moment to ask whether the platform still fits the workload, which is the question our AI operating model consulting frames around ownership, cost and control rather than vendor demos.

Three tests settle most cases. The first is the region test: can you say, today, which cluster your workspace will sit in after the retirement and why. The second is the action test: how many actions your agents call, who owns each target system, and what it would take to rebuild them elsewhere. The third is the commitment test: whether your call volume justifies Synthflow's published enterprise entry tier, which our pricing comparison and the voice AI buyer's guide linked above set out with sources. A team that answers all three comfortably should probably stay, and a team that cannot is ready for a structured placement review.

How do you move off Synthflow without dropping calls?

A team moves off Synthflow without dropping calls by sequencing four steps: confirm the data region decision in writing, rebuild the custom actions on the new platform, repoint the Twilio or SIP numbers for a small share of traffic, then run both platforms side by side until outcomes match. Where UK numbers sit in the customer's own Twilio or carrier account, no number port is needed and the old agents remain a rollback path.

Step one is the region decision, recorded whether you stay or go. If you are leaving, you still need the old workspace to behave predictably until cutover, so confirm with Synthflow which cluster it will sit in and when. Step two is the action layer. Export every agent version you rely on, list each custom action with its target system and authentication, and rebuild them on the new platform before any caller reaches it. This is the step teams underestimate, because the version export makes the agent look portable, and it is where our vendor exit guide checklist earns its keep.

Step three is the number move. With numbers in your own Twilio account or SIP trunk, which you should confirm first, point a small set of numbers, or a share of inbound routing, at the new platform. Step four is shadowing: run the same test scenarios through both platforms, compare call outcomes and transfers to humans, and widen traffic only when the results match. Keep the old agents published and the old routing documented, so reverting is a configuration change. Our release management and change control guide covers rollback discipline in more depth, and the cutover itself deserves a named owner and a written rollback trigger.

The same diagnostic logic underpins our supervised AI execution office, a retained team that supervises delivery alongside the client's operations leads.

What is the best Synthflow alternative for UK teams in 2026?

The best Synthflow alternative for a UK team in 2026 depends on who maintains the agents and where the data must sit. A developer-led team may prefer an orchestration platform it configures in code. An operations-led team that wants per-country call compliance rules shipped by default may prefer a managed platform such as Dilr Voice. An agency reselling under its own brand may find no alternative beats Synthflow.

Name the criteria before naming vendors. For a UK buyer they are: where call data is stored and processed, and whether that can be chosen per workload; whether numbers stay in the customer's own Twilio or carrier account; how custom actions are built and versioned; what call-compliance controls ship by default; and how quickly a team can test before committing. Vapi and Retell AI are developer-first options on that path, and our orchestration versus platform guide explains that split. The buyer's guide linked in the introduction carries the full matrix with sourced figures.

Here is where Dilr Voice fits, using only what its live pages state. It is a multi-agent platform that chains a greeter, qualifier, knowledge, customer care and action agent into one call, each with its own prompt, model and tools. It ships per-country compliance rules by default, including recording consent, do-not-call checks, opt-out recognition, permitted calling hours and an audit trail on every call. It runs on Google Cloud with encryption at rest and in transit, and offers dedicated tenancy and regional data residency options for enterprise. Response time is under 500 milliseconds on DILR's own measurement, and it supports 30+ languages.

The concession matters as much as the claim. Dilr Voice is a newer brand with a thin public review footprint, it does not offer agency white-label reselling, and it shows no public certification. A team that needs a vendor certification report before procurement should ask every shortlisted vendor for the report itself, not a badge. You can test Dilr Voice with a free trial, with no credit card required, and read the wider decision frame in our enterprise AI voice agents guide.

How should you judge a Synthflow move on return, not features?

A Synthflow move should be judged on payback: the one-off cost of rebuilding actions, re-papering transfers and re-running tests, set against the change in running cost and in calls resolved without a human. If the move only swaps one feature list for another, it rarely pays back. If it fixes a data region problem the business must solve anyway, the payback case is already half made.

Our AI voice ROI framework sets out the method: baseline the current cost per handled call, model the one-off transition effort honestly, and measure resolved calls rather than answered ones. For a Synthflow workload, the transition effort is dominated by the action rebuild and the shadow period, both of which scale with the number of integrations rather than the number of agents. Count your actions first; that number predicts the project better than any vendor estimate, and it is the first figure worth putting in front of a steering group.

Does the Synthflow cluster retirement affect every customer?

Not as written. Synthflow's migration notes say they apply to enterprise customers still running on the legacy Global cluster, and its region page treats a US or EU workspace as permanent in that region. Synthflow says nothing moves without notice and that region confirmation emails go out through October, so a team unsure of its status should check the Customer Region setting in its workspace preferences.

If that check shows the legacy cluster and no region has been confirmed, treat it as this week's job. Our Dilr Voice FAQ answers the common questions about our own platform.

Can a Synthflow workspace hold both US and EU data?

No. Synthflow's documentation says one workspace cannot hold both regions, and a second region needs a second workspace, provisioned and billed separately. A UK group with some workloads that must stay in Europe and others that may sit in the US would therefore run two workspaces, with separate agents, actions and API base URLs, and should plan its records of processing on that basis.

Who is responsible for AI disclosure on a Synthflow call?

The business deploying the agent is. Synthflow's AI transparency statement says customers deploy and operate the agents they build, and that configuring AI disclosure in line with the rules for their use case and jurisdiction is the customer's responsibility. Synthflow provides configurable greeting and consent messages, but choosing the wording and making sure it plays stays with the UK business deploying the agent.

The statement itself is on Synthflow's AI transparency page. Check the equivalent wording for any platform on your shortlist; in our experience the disclosure controls ship with the software and the duty sits with the business that deploys the agent. Our voice AI category hub collects the wider compliance reading, and the Dilr Voice product page sets out the call-compliance controls that ship by default.

Want to see this in production? Try Dilr Voice live, read the voice AI buyer's guide, talk to our AI execution office, or read about our approach to placing AI inside enterprise systems.

Written by the Dilr.ai engineering team, practitioners who ship enterprise AI in production. Follow us on LinkedIn for shipping notes, or subscribe via the RSS feed.