Sherlock FAQs

1. How does Sherlock integrate with Zoom and Microsoft Teams?
Sherlock integrates with Zoom and Microsoft Teams using calendar-based triggers and meeting metadata. It connects via calendar plugins or API hooks that detect scheduled interviews. Once triggered, Sherlock silently joins the meeting or activates locally through its desktop agent, requiring no change to the existing video conferencing flow.

2. Does Sherlock work in live interviews or only recorded ones?
Sherlock currently works only in live interviews. It provides real-time monitoring and flags integrity issues such as whispering, screen sharing, and off-camera assistance during the session.

3. Does it require candidate-side installation or run in the browser?
For full protection and agentic performance, Sherlock requires candidates to install a lightweight desktop application. This allows it to monitor screen activity, mic input, tab switching, and more. Interviewers do not need to install anything.

4. Can it run passively without interrupting the interview experience?
Yes. Sherlock operates invisibly in the background. It collects data without interfering with the candidate or interviewer, preserving a natural interview experience.

5. Can Sherlock be embedded into our interview scheduling workflow?
Yes. Sherlock can be integrated into scheduling tools like Google Calendar, Outlook, Calendly, or ATS platforms. Once set up, it auto-activates during scheduled interviews, removing the need for manual setup.

6. Does Sherlock integrate with applicant tracking systems (ATS)?
Yes. Sherlock supports integrations with Greenhouse, Lever, Workday, iCIMS, and other ATS platforms via APIs. This enables automatic syncing of interview metadata and Sherlock-generated reports.

7. Where is the data stored, and is it encrypted?
Sherlock uses enterprise-grade security with AES-256 encryption for data at rest and TLS 1.2+ for data in transit. Data is stored on SOC 2 and ISO-certified infrastructure hosted on AWS.

8. Do you store candidate PII, and how is it protected?
Only essential PII is stored—typically name, email, and interview ID. All data is encrypted, governed by RBAC, and logged for auditing. Anonymization features are also available.

9. Is Sherlock compliant with industry regulations and standards?
Yes. Sherlock complies with GDPR, HIPAA, SOC 2, ISO/IEC 27001, and CCPA. It also supports region-specific configurations and customer-specific consent flows.

10. Can it be hosted in a private cloud or on-premise?
Yes. Sherlock can be deployed in a dedicated VPC, hybrid environment, or fully on-premise. This is ideal for organizations with strict data residency or internal compliance policies.

11. How do we handle candidate consent and privacy notifications?
Sherlock includes customizable consent flows. Candidates see a clear explanation before the interview and must opt-in. All consent is logged and auditable to meet compliance needs.

12. What types of violations does Sherlock detect?
Sherlock detects over 20 types of behavioral and technical violations, including:

12. How accurate is Sherlock, and what is the false positive rate?
Sherlock’s detection algorithms are trained using thousands of hours of real interview data and verified incidents. It uses multi-modal inputs—video, audio, and behavioral logs—to triangulate events. The current false positive rate is below 2%, and every alert is accompanied by a confidence score and explanation, allowing reviewers to make final judgment calls.

13. Does it offer real-time alerts or only post-interview summaries?
Sherlock provides both. During a live interview, Sherlock can issue real-time alerts via Slack, email, or its monitoring dashboard. Afterward, it generates a complete integrity report including evidence, violation breakdowns, and a candidate trust score.

14. Can Sherlock detect if a candidate is reading from a script or getting external help?
Yes. Sherlock uses gaze tracking, behavioral fluency analysis, and language similarity models to detect scripted or AI-generated responses. It flags abnormal reading patterns, long pauses before responses, and ChatGPT-like language structures.

15. Can Sherlock detect whispering, screen sharing, or use of AI tools like ChatGPT during interviews?
Yes. Sherlock detects whispering by analyzing audio signal intensity and frequency. It flags screen sharing or mirrored displays and uses an LLM signature-matching model to identify AI-generated responses like those from ChatGPT.

16. Does Sherlock detect multiple people on camera or speaking in the background?
Yes. Sherlock uses facial detection and audio source separation to identify multiple faces or voices. If another person appears on screen or speaks in the background, it generates a violation with supporting evidence.

17. What kind of reports does Sherlock provide?
Sherlock provides reports that include:\n- A summarized integrity score

18. Can these reports be shared or exported?
Yes. Reports can be exported as PDFs, shared via secure dashboard links, or integrated into ATS/HRIS systems. Sherlock supports RBAC for access control and audit-compliant sharing.

19. Can we adjust violation thresholds or customize detection logic?
Absolutely. Sherlock allows clients to define what counts as a violation, set severity thresholds, and enable or disable specific modules. This ensures the system reflects your internal hiring and compliance policies.

20. Can interviewers or reviewers leave notes and comments in the report?
Yes. Interviewers and reviewers can leave time-stamped comments, add tags, and annotate the report. This promotes team-based hiring decisions and transparent documentation.

21. Is there an audit trail for compliance and legal reviews?
Yes. Every viewer interaction, comment, export, and flag resolution is logged in a tamper-proof audit trail. This is essential for regulated environments like healthcare, finance, or government hiring.

22. Can we run a trial or proof of concept?
Yes. Sherlock offers a 2–4 week proof-of-concept (PoC) engagement where clients can test the platform with real interviews. This includes full access to all detection features, integrity reports, and dashboards. At the end of the PoC, clients receive a summary of detection accuracy, user experience feedback, and recommendations based on observed data.

23. Do you offer customer references or case studies?
Yes. We provide anonymized case studies across industries like healthcare, technology, and finance. These include metrics such as improvement in quality-of-hire, reduction in ghosting, and bias elimination. Direct customer references can also be arranged upon request for enterprise clients.

24. What’s the pricing model?
Sherlock offers flexible pricing:

25. What’s the ROI of Sherlock compared to traditional methods?
Sherlock drives ROI by:

26. Will interviewers need special training to use Sherlock?
No special training is required. Interviewers conduct interviews as usual. Sherlock works silently in the background and auto-generates reports. A 30-minute onboarding session is sufficient for recruiters and hiring managers.

27. How do we handle candidate consent and privacy notifications?
Sherlock has built-in, customizable consent workflows. Before the session begins, candidates are shown what Sherlock monitors and must opt in. This consent is time-stamped and stored for audit purposes. Organizations can embed their legal and privacy notices within this flow.

28. How does Sherlock handle edge cases like poor internet or neurodiverse candidates?
Sherlock flags such conditions but does not penalize them by default. All flagged events are reviewed in context by hiring teams. We also offer detection model calibration to reduce bias and improve fairness for special candidate populations.

29. Can Sherlock be enabled selectively by department or role?
Yes. Sherlock can be scoped to run only for certain roles, departments, geographies, or interview types. Admins can control settings through the dashboard or API-level integration.

30. Is there a way to provide candidates with transparency about how they’re being evaluated?
Yes. We offer a candidate guide explaining what Sherlock tracks, why it exists, and how their data will be used. This promotes transparency and helps candidates feel psychologically safe, especially in sensitive or high-stakes scenarios.