Effective 1 September 2026
Privacy policy and notice at collection
This notice explains how Sansa Group AB operates ResearchBuddy, what data is processed, where it goes, and how you control it.
Controller and contact
ResearchBuddy is operated by Sansa Group AB, trading as SansaVision, at Soldathemsgatan 20, SE-415 28 Göteborg, Sweden. For privacy matters email support@researchbuddy.com or use the privacy-rights form. Messages are delivered to support@sansavision.com.
Data categories, purposes, and retention
| Category | Examples | Purpose | Retention |
|---|---|---|---|
| Publication pathway and integrity records | Paper type, field, goals, milestones, readiness gates, ethics and data status, contributor roles, AI-use disclosures, expert reports, author responses and submission-package state. | Organize work from direction to submission, preserve provenance, support human review, and create user-directed disclosures and exports. | Follows the project lifecycle; integrity, authorship, transaction, dispute, and audit records may remain where necessary for scholarly accountability, legal claims, or compliance. |
| Opportunity and matching records | Field, paper type, constraints, sourced opportunity suggestions, saves, dismissals, reviewer searches, open review calls, applications, expertise, availability, credentials and conflict screening. | Suggest feasible research directions and match researchers to relevant, outcome-independent expert services without promising novelty or publication. | Opportunity jobs and suggestions remain with the account until deleted or converted; matching and conflict records follow the applicable profile or engagement lifecycle. |
| Expert marketplace and payout records | Service scope, deliverables, confidentiality, price, platform fee, verdict, payout onboarding status, payment identifiers, refunds, disputes and transaction ledger. | Form and perform expert engagements, prevent conflicts and coercion, fund work, pay experts, reconcile fees, handle support, tax, fraud, refunds and disputes. | Engagement and transaction records are generally retained for seven years or longer where required for tax, payment disputes, fraud prevention, professional accountability, or legal claims. ResearchBuddy does not receive full card or bank-account numbers. |
| Institutional program records | Organization, cohort, enrollment, supervisor assignment, program policy, review budget, pathway association and aggregate milestone progress. | Operate supervised publication programs and show authorized people relevant progress without automatically exposing private project content. | Follows the program and organization lifecycle, subject to education, employment, contract, audit, research-integrity, and legal retention requirements. |
| Feed personalization activity | Post impressions, coarse dwell-time buckets, saves, follows, discussions, source opens, hides, topic preferences and recommendation explanations. | Rank a diverse research feed, explain why an item was suggested, reduce repetition and honor show-less or mute choices. | Raw feed events are normally deleted after 90 days. Aggregated topic preferences remain with the account until you reset recommendations or delete eligible account data. |
| Public profile and community activity | Profile image, name, unique public address, address-change requests, role, affiliation, fields, methods, public posts, reactions, follows, collaborator relationships and reports. | Make researchers and public work discoverable, support credible discussion, recommend relevant people and topics, and moderate misuse. | Visible until you hide or delete eligible content; moderation and integrity records may remain where necessary to protect the community or establish legal claims. |
| Private profile materials | A CV, extracted profile suggestions, collaboration preferences, availability and section-level visibility choices. | Suggest expertise and work-history fields, improve collaborator matching, and apply your privacy choices. | The original CV remains private unless you explicitly publish it and is deleted with the eligible account or earlier at your request; extracted suggestions are not published until you approve them. |
| Identifiers and account data | Name, email, account and session identifiers, organization membership, authentication and security events. | Account access, service delivery, fraud prevention, support and legal compliance. | Account life; active sessions up to 30 days; security records normally 12 months. |
| Research and collaboration content | Uploaded files, notes, annotations, claims, evidence links, tasks, discussions, document history and publication choices. | Store, retrieve, collaborate, cite, export, publish when instructed, and provide requested AI tools. | Until the project owner deletes it; operational backups are removed on a rolling basis, normally within 30 days. |
| Research-assistance and search records | Questions, authorized evidence excerpts, source IDs, outputs, review status and integrity records. | Prepare source-backed responses, prevent misuse and preserve an inspectable research history. | Run metadata and accepted work follow the project; temporary processing follows our provider agreements. |
| Commercial records | Plan, credit ledger, payment and subscription identifiers, transaction state and support history. | Checkout, entitlements, refunds, reconciliation, fraud prevention, tax and accounting. | Transaction and tax records are generally retained for seven years; card numbers are not received by ResearchBuddy. |
| Consent and communications | Cookie choice, GPC status, legal-policy acknowledgement, support messages and privacy requests. | Honor preferences, prove compliance and respond to requests. | Cookie choices are renewed at least annually; support/privacy records normally three years after closure; legal records as required for claims and compliance. |
| Messages, invitations and notification preferences | Direct messages, Research Rooms, project channels, reads, mentions, invitations, delivery status, unsubscribe choices and digest timing. | Deliver collaboration, preserve context and consent, notify you according to your settings, prevent unwanted contact and maintain an auditable invitation history. | Messages follow the relevant conversation or project lifecycle; delivery and preference records are retained while the account is active and as needed for suppression and legal compliance. |
Sources and required data
Most data comes from you, your collaborators, your institution, the files and public-source queries you submit, authentication and security signals, and payment events. Account identifiers are required to provide the service. Optional analytics is not required. If you do not provide necessary account or project data, the corresponding feature cannot operate.
Legal bases in Europe
- Contract: account access, storage, collaboration, requested AI/search operations, exports, billing and support.
- Legitimate interests: service security, abuse prevention, reliability, limited operational measurement and defending legal claims, balanced against user rights.
- Consent: optional analytics and any other feature explicitly presented as optional consent; consent can be withdrawn as easily as it is given.
- Legal obligation: tax, accounting, regulatory requests and legally required records.
Research assistance and public-source search
Project authorization is checked before evidence is retrieved, and only evidence selected for the request is included. When you enable public-source search, your question is sent to scholarly discovery services and public candidate pages may be opened for context. Private project excerpts are not sent as public-search queries. Deeper review may involve a contracted inference provider under our data-protection terms.
Profiles, recommendations, and daily research suggestions
A basic profile becomes public after onboarding unless you hide it. Optional profile sections, your CV, private projects, saved items and messages remain private unless you deliberately change their visibility. We use fields, follows, topic feedback, freshness and evidence-quality signals to recommend researchers, posts, questions and papers. Follower totals and raw upvotes are not treated as scholarly authority. Suggested daily questions are labeled, and paper suggestions link to a real scholarly source. These recommendations do not create legally significant decisions. Feed impressions and coarse dwell-time buckets help reduce repetition and improve relevance. You can hide an item, request fewer similar items, inspect the recommendation reason, or reset behavioral recommendation history from Community.
Expert matching, conflicts, and institutional oversight
Expert discovery uses fields, methods, services, availability, language, turnaround, and verification status. Compensation never depends on the review verdict. Conflict declarations are shared only as needed to decide whether an engagement can proceed and to preserve its integrity history. Institutional program managers see enrollment, assignment, and milestone information permitted by their organization role; joining a program does not automatically expose private project content. Payment providers independently collect identity, bank, tax, and eligibility information for expert payouts. Open review listings expose the scope, field, fee, timing, and chosen disclosure level. Application notes are visible to the listing owner; private manuscripts remain unavailable until selection and conflict screening. Review records distinguish human review from human-led review that used assisted checking tools.
Community safety and moderation
Reports, blocks, mutes and limited activity context may be reviewed to enforce our community rules, protect researchers and investigate abuse. We do not create synthetic researcher profiles, comments or engagement. Moderation decisions are recorded with an immutable audit event where required, and you can appeal through support or the privacy-rights form when personal data or automated assistance materially affected the decision.
Recipients and service providers
Contracted providers support hosting, security, storage, research assistance, public-source discovery, email, payments and operational processing. Payment providers independently process payment, tax, receipt, refund, subscription and transaction-support data, with the checkout provider acting as merchant of record where displayed. Authorized collaborators receive only project data permitted by their role. We do not sell personal information or share it for cross-context behavioral advertising. You may request our current subprocessor list through the privacy-rights form.
International transfers
Data may be processed outside your country where these providers operate. We use available adequacy decisions, contractual safeguards and provider transfer mechanisms as applicable. You may request information about the safeguard relevant to your data through the privacy-rights form.
California disclosures
During the preceding 12 months, the service may have collected the categories described above: identifiers, customer records, commercial information, internet or service activity, professional or education information supplied in research content, and inferences represented by AI artifacts. ResearchBuddy does not sell these categories and does not share them for cross-context behavioral advertising. It does not use sensitive personal information to infer characteristics or for purposes requiring a right-to-limit link. A Global Privacy Control signal disables optional analytics.
Your choices and rights
You can export account and project data, delete an eligible account, reopen Cookie choices from every footer, and submit access, correction, deletion, portability, restriction, objection, opt-out, limitation or appeal requests. Details and statutory timelines are in the GDPR and California rights matrix. You may also complain to the Swedish Authority for Privacy Protection (IMY) or the regulator where you live.
Automated decisions and children
ResearchBuddy AI produces advisory artifacts but does not make solely automated decisions with legal or similarly significant effects about users. Human review remains required. The service is not directed to children under 16, and a school or guardian must provide any authorization required by applicable law.
Security and changes
We use access checks, encryption in transit, private object storage, signed webhooks, rate limits, security logging and recovery-oriented exports. No system is perfectly secure. Material policy changes will be dated and presented before a new legal acknowledgement or consent is requested where required.