Information we collect
Account and contact information, work experience, education, skills, job preferences, salary expectations, availability, CVs, supporting documents, dedicated Gmail credentials, and Naukri credentials are collected through the candidate workspace. Application progress and administrative notes may be recorded by the team. An interest form separately collects your name, email, and chosen interest.
Why account credentials are requested
The service requires access to the job-search accounts you authorize so the team can handle applications offline. A fresh, dedicated Gmail account is recommended rather than your primary personal mailbox. The Gmail ID and password and Naukri login details are submitted through the authenticated credential vault. Sharing them gives the authorized team access to those accounts and the information they contain.
Credential storage and administrator access
Job-search passwords are stored encrypted rather than as ordinary profile fields. Authorized administrators can access credentials for the application service, with scoped access and audit records. Passwords must not appear in public pages, analytics, ordinary logs, or profile exports. They are not sent to an application agent; no autonomous application execution is enabled in this beta.
How your information is used
The team uses your submitted information to understand your experience, target suitable roles, prepare applications, and complete them offline. An application intentionally shares relevant candidate information with the employer or job site. The team may contact you to clarify missing information. A received beta-interest request is not treated as a verified email address.
Optional AI profile import
When you choose Extract with AI, PDF.js first extracts text from a selected PDF. Scanned or text-sparse PDFs use a private Docling OCR function in AWS Mumbai with offline models. The extracted text, pasted text, or fetched public webpage content is sent to Claude Sonnet 4.6 on Amazon Bedrock to suggest professional profile fields. Processing may occur outside India. The model runs through a global inference profile; source content and model responses are not logged by the application. Stored Gmail and Naukri credentials are not included. Source files are processed for this request without being added to document storage by the importer. Review and correct suggestions before applying them; existing profile values are preserved. LinkedIn and other pages that require sign-in may need a profile PDF or pasted text instead.
LinkedIn sign-in and optional profile import
LinkedIn can be used to create or access a candidate account. We use its stable account identifier and available verified email for authentication. A sign-in connection is stored with the account; LinkedIn passwords are never collected. Existing accounts must be verified before linking and are not merged by email. Separately, Import my details from LinkedIn asks for explicit consent before fetching name and available email for a five-minute encrypted review preview. Only reviewed details fill empty profile fields. Full experience, education, skills, languages and certifications can be imported from supported files in a LinkedIn data export, with a separate consent step; messages and connections are excluded. Archive files are processed for the request without being retained as documents or sent to AI. Provider access tokens are used for the request and discarded. Sign-in or profile import does not authorise job applications or sharing Gmail and Naukri credentials.
Service providers and processing locations
Mumbai is the primary AWS data-region direction. Frontend hosting, account authentication, email delivery, and the job sites used for applications may process information elsewhere. Primary storage in India does not imply that every processing step remains in India. Service-provider access should be limited to what is required for the configured functions.
Current service and future Saaya offering
The current beta does not run application browser agents or live Saaya interview practice. It does not grant practice minutes, take recharge payments, or create live practice transcripts or assessments. The future practice offering will need its own session authorization, context isolation, usage accounting, processor disclosure, and consent before activation.
Your controls and requests
Review and update the profile, preferences, and documents in your account. Use the privacy controls to request an export or deletion. Ordinary exports must exclude decrypted credentials. Account deletion should revoke service access and remove candidate data under the applicable policy, while restricted audit records may be retained where required. A change made online may need clarification with the team if offline work is already in progress.
Cookies and local storage
Account authentication uses the storage necessary to maintain access and protect sessions. The public marketing website does not require advertising cookies. Optional analytics or recording should not be activated without the relevant notice and choice. Integrated providers may have their own service behavior and policies.
Questions and changes
Use your authenticated account’s privacy controls for data requests. You may also reply to a beta invitation or service communication from the team. Material changes to data use or new offerings should be explained before the changed use begins.