Vela

Vela acts in connected calendars, sheets, boards and messaging channels. The staffing case study says Vela "connects to the client's Google Sheet" where interviewers post availability "and the Monday.com board" where candidates move through the pipeline, calls candidates by phone, emails and texts them, creates calendar invites, and takes Slack instructions.

The privacy notice says Vela accesses Google Calendar and Gmail data through the Google API Services and connects Google or Microsoft calendars, with revocable access, and the home page names email (Gmail and Outlook), SMS, WhatsApp, Slack, Teams and phone. Vela creates invites, places calls, sends messages, and reads and updates connected boards.

No integrations directory is published, and the only credential scoping described is revocable calendar access.

Sourcetryvela.ai/case-studies/scaling-interview-ops, /privacy, home; readread 2026-09-16

Scheduling runs as a branching workflow with retries across channels, configured in natural language.

In the staffing case study, on a pipeline change Vela "pulls interviewer availability from the connected Google Sheet, calls the candidate by phone first, if they do not answer, it follows up by email", books a slot, sends double-blind calendar invites, sends reminders 24 and 2 hours before, and on cancellation "automatically attempts to reschedule" by phone, email or SMS, continuing on whichever channel the candidate replies.

The customer configures it in natural language ("First-round interviews are 30 minutes. In-person interviews are 60 minutes."), and a new client or role is set up in about 15 minutes. The phone-then-email branch, channel routing and repeated reschedule attempts make fallbacks and retries first-class, and one system runs across clients.

Versioning and debugging are not documented, and the workflow is described in a customer story rather than product documentation.

Sourcetryvela.ai/case-studies/scaling-interview-ops, home; readread 2026-09-16

Scheduling decisions draw on the thread, calendars and sheets Vela is connected to, assembled at run time. The home page says Vela reads context from the thread it is copied on and reads your calendar; the staffing case study says it "reads interviewer availability directly from the source of truth" (a connected Google Sheet) and matches candidates to open slots by time zone, duration and location; and the privacy notice says Vela sees only the conversations and calendars a user brings it into.

No maintained index, knowledge base or citation surface is documented.

Sourcetryvela.ai/case-studies/scaling-interview-ops, home, /privacy; readread 2026-09-16

Vela escalates to a person after repeated failures, and holds a permission step by default that the customer can lift per action.

The staffing case study says that "after multiple failed attempts, it escalates to a human on the team", and that customers set behavior in natural-language instructions such as "On rejection, immediately call the candidate back to reschedule without asking me for permission", which shows a permission step that applies by default and is waived per action type.

The terms say "AI-generated suggestions can be wrong; you remain responsible for reviewing outputs and final decisions (e.g., sending a message or confirming a time)". A documented escalation rule plus a default permission step the customer can lift for a named action is a configurable hold, with autonomy set per action through instructions.

The default permission step is shown through a customer's instruction rather than documented directly, and routing approvals to other channels is not described.

Sourcetryvela.ai/case-studies/scaling-interview-ops, /terms; readread 2026-09-16

Encryption and a no-training commitment from its AI providers are documented, but no attestation and no customer access controls.

The privacy notice says Vela maintains "administrative, technical, and organizational safeguards... including encryption in transit and at rest", that its "AI providers are contractually prohibited from training their models on your content", that Google user data follows the Google API Services Limited Use requirements and is read by people only in listed cases, and that calendar access can be revoked at any time; the terms say integration data travels over encrypted channels, with encryption at rest required of providers. That data handling and training policy, with encryption, is a first party security posture.

No certification or report is published, and no customer facing access model (roles, SSO, SAML) is documented. The privacy notice links a trust center at tryvela.ai/trust, which returned "Page Not Found" (HTTP 404) on 16 September 2026.

Sourcetryvela.ai/privacy, /terms, /trust (404); readread 2026-09-16

Operations teams get one dashboard over the scheduling work and a measure of time saved. Vela's staffing case study says the operations team "went from monitoring 17 systems multiple times a day to checking one dashboard", and the home page says "Vela measures actual time saved once deployed"; reminders and confirmations go to all parties.

Step by step inspection of what Vela did, a per meeting activity history and an audit log of what it sent and changed are not documented.

Sourcetryvela.ai/case-studies/scaling-interview-ops, home; readread 2026-09-16

A scheduling conversation carries across channels. In the staffing case study, if a candidate replies to a confirmation email or calls back the number that contacted them, "Vela continues the conversation" across channels; customers write standing natural-language instructions per workflow; and the privacy notice lists scheduling preferences among information collected.

No memory layer with a stated scope, lifetime, or way to review, edit or delete memories is documented; per-workflow instructions are configuration. The home FAQ asks "Can I set preferences for meeting types?", but the answer is not in the page text.

Sourcetryvela.ai/case-studies/scaling-interview-ops, /privacy; readread 2026-09-16

No hosting region or residency choice is stated for this cloud service. The privacy notice says Vela is headquartered in the United States and "may process personal information in the United States and other countries", with Standard Contractual Clauses where required; no hosting region, residency choice or customer hosted option is published. Processing in the US "and other countries" names no location a customer can rely on or choose.

Sourcetryvela.ai/privacy, /terms; readread 2026-09-16

One scheduling assistant, configured per client, is the whole offer. The home page sells "One assistant. All your meetings. End-to-end.", and the staffing case study describes each client's workflow being set up by the customer in natural-language instructions in about 15 minutes. No catalog, template library or set of separately adoptable agents is published, and the sitemap lists only three case studies. A single agent configured per client offers no packaged unit to adopt.

Sourcetryvela.ai home, /case-studies/scaling-interview-ops, /sitemap.xml; readread 2026-09-16

Work reaches Vela without anyone asking, from pipeline changes and calendar conflicts, as well as through the channels people use to reach it. Vela's staffing case study says that when a candidate is moved to the interview stage on Monday.com, "Vela detects the change automatically.

No manual trigger needed", and that when a candidate cancels, Vela "automatically attempts to reschedule... without human intervention"; the home page says Vela "monitors your availability in real-time and reschedules automatically if conflicts arise" and follows up automatically when someone doesn't respond. Users also start work by CCing vela@yourdomain.com on email, SMS or WhatsApp, texting Vela, or messaging it on Slack. A pipeline stage change and a calendar conflict both start work with nobody asking, so triggers are both event-driven and conversational.

Duplicate-trigger handling is not described.

Sourcetryvela.ai/case-studies/scaling-interview-ops, home; readread 2026-09-16

The service runs on unnamed third party AI models, with no choice offered. The privacy notice says "we use third-party AI model providers to power features of the service", under terms that bar training on customer data, and lists AI model providers among service providers; no model or provider is named, no routing between models is described, and no model choice is offered to the customer or an admin. A generic reference to unnamed providers discloses neither routing nor choice.

Sourcetryvela.ai/privacy; readread 2026-09-16

Whether Vela offers an API is not public. The home page FAQ asks "Is there an API?", but its answer is collapsed and does not load as page text without a browser. No developer, docs or API page is linked from the site, and the sitemap lists only three case studies.

Sourcetryvela.ai home FAQ; triedread 2026-09-16

What Vela measures is the customer's outcome, not a change to its own behavior before it goes live. The home page says "Vela measures actual time saved once deployed", an outcome metric for the customer's scheduling load rather than a test of the agent; the Scheduling Benchmark is a published study of scheduling emails, not a tool customers run; and the case study describes natural-language setup with no test step. No simulation, test run or scored evaluation of Vela's behavior is documented.

Sourcetryvela.ai home, /case-studies/scaling-interview-ops; readread 2026-09-16

Rather than operating screens, the assistant works through email, SMS, WhatsApp, Slack, Teams and phone calls, connected calendars, and connected Google Sheets and Monday.com boards. No browser, desktop or screen operation is described.

Sourcetryvela.ai home, /privacy, /case-studies/scaling-interview-ops; readread 2026-09-16