Frequently asked questions

In the cloud version, data is stored on AWS servers located in the US or the EU. Raw audio and video recordings are encrypted on AWS S3 using AES-256, and data in transit is protected with TLS 1.2 and above. Each customer’s data is partitioned with a unique Tenant ID in DynamoDB so it is fully isolated, and encryption keys are managed through AWS KMS. On request, data can be stored exclusively in the European region. Sensitive information and trade secrets are safeguarded through an NDA, a Data Processing Agreement (DPA), and a Trade Secret agreement, all signed before any data is processed. No content is used, shared, or retained outside the scope of processing.

No. We do not have access to your data, and it is never used for model training. Access is provided only at your explicit request — for example, when a specific meeting needs to be restored or an ongoing technical issue needs to be investigated.

Spiky maintains SOC 2 Type II (audited quarterly), ISO 27001-equivalent processes, and GDPR/KVKK compliance. Identity management uses AWS Cognito, RBAC, mandatory MFA, and Azure SSO (LDAP is not used). With Zero-Day Retention, recordings and transcripts are not kept beyond the processing window, and deletion windows are configurable (immediate, 6, 24, 72 hours, or custom). Technical deletion disables the AWS KMS key, triggers S3 lifecycle deletion, and overwrites the data — verifiable via Vanta, AWS CloudTrail, and SOC 2 logs. Spiky falls outside the EU AI Act’s prohibited and high-risk categories, and emotion/behavioral analysis can be disabled on request.

An automatic notice is sent before the meeting, and if a participant leaves within the first four minutes, no recording is kept. The process is fully compliant with GDPR and KVKK.

Yes, we offer and support on-premise deployment. The architecture is containerized and Kubernetes-compatible, with two layers: a recording and bot layer that uses only CPU (no GPU required), and a processing and LLM layer for transcription, summaries, and insights that can use your organization’s own model services. In an on-premise deployment, all real-time streaming stays within your own environment; no data leaves the Spiky components toward the cloud (the only outbound connection is the bot joining the meeting platform). Pricing is shared as a custom offer tailored to your requirements after a technical scoping. Learn more on our on-premise deployment page.

Diarization depends on how the meeting is recorded. If a bot joins the meeting, it automatically detects the names of the online participants. For in-person or hybrid meetings, if you record with our iOS app or Android app, or upload a voice recording, we separate speakers via diarization and initially label them "Member 1", "Member 2", etc. (a beta capability); if a participant says their name, the system can also assign the real name. Current limitation: when several people (e.g., five) speak through a single microphone in a physical room, we can’t yet separate those individual speakers — we’re actively working on this. Your sales representative can share more details.

Microsoft Copilot serves a similar purpose on paper, but there are a few key differences. (1) Turkish accuracy: existing note-taking tools often underperform in Turkish, whereas Spiky’s Turkish accuracy is notably high. (2) Coverage: Copilot only captures meetings inside the Microsoft ecosystem (e.g., Teams); if a customer sends you a Zoom or Google Meet link, it can’t record those — Spiky captures meetings across all conferencing tools. (3) Added capabilities: use cases Copilot doesn’t offer, such as Spiky Signals (revenue signals and reporting) and Spiky Whisper (real-time live coaching). Spiky also closes the cross-functional alignment and task-ownership gap with clearly assigned action items.

We integrate natively with Microsoft Teams and Outlook Calendar. For other software such as Microsoft Dynamics, integration can be provided through Zapier or n8n. In addition, Azure SSO is supported, and Active Directory integration is currently under development.

Yes. As long as each session is recorded individually, Whisper works alongside any dialer application — including Dialpad, RingCentral, Aircall, Orum, Nooks, JustCall, 3CX, and Five9. See all integrations. The Whisper desktop app can also capture your computer’s system audio and microphone.

Yes. You can connect Spiky to 6,000+ apps with Zapier and trigger automations the moment a meeting analysis completes — no code required. See the Zapier integration page for details.

Yes. With the verified Spiky n8n node you can build low-code automations, self-hosted or in the cloud. See the n8n integration page for details.

Yes. With Spiky MCP you can connect Spiky to Claude, ChatGPT, and other AI agents, exposing your conversation data and revenue signals to AI. See the MCP integration page for details.

The Spiky Platform API is available and lets you push recordings and pull analysis and signals programmatically. See the API integration page and the API guide for details. There is currently no Whisper API.

Signals can ingest data from tools like Gong, Chorus, and Fireflies, from dialer recordings, and from your existing Spiky Pulse account. Signals digests your existing call and phone recordings at the end of the day and turns them into revenue signals.

Yes. For enterprise and paid customers, our team can carry out this upload. Data can be imported from sources such as Notion.

Yes. Using Workflows, an automatic notification can be sent when a call or meeting score falls below a defined threshold.

Yes. A scorecard can be generated automatically after each meeting; this requires enabling the Spiky Agent feature.

Reports can be sent daily, weekly, or monthly. Email notifications can be set up through Workflows.

Turkish is positioned as a native-level differentiator with around 95% accuracy* and can be improved further with a custom dictionary. (*based on our internal tests)

The Enterprise plan has features similar to the Premium plan. Its main differences are an SLA (service level agreement), priority support, and API access.

Signals pricing depends on factors such as the number of users and reporting frequency, and is offered with an annual commitment (payable monthly). Your sales representative can share a tailored quote.

Yes, it’s possible — Whisper and Pulse Platform can be purchased separately. The right setup depends on your use case, though, so we recommend talking to your sales representative. (You’ll access Pulse Platform to configure Whisper.)

No. Under our fair-usage policy, a single license cannot be shared across a team or organization; each user needs their own license. This is consistent with comparable solutions.

Yes. We offer dedicated onboarding sessions with our Customer Success Managers (CSMs).

For the Pulse Platform, a POC typically runs 2–4 weeks depending on the complexity of the project; in most cases, one week is enough to test language accuracy and get a feel for the platform. For the Signals Platform, timelines vary — your sales representative can provide more detail.

We offer both monthly and annual contracts. At this time, there is no minimum-user requirement.

As a startup, all feature requests are ranked and compared against other requests based on whether comparable companies would also benefit, the difficulty of implementation, and scalability.