Grounded on your knowledge base
Ingest your policies, handbooks, JDs, SOPs and playbooks so every answer is anchored to your approved, current documents - with citations back to the source paragraph.
Your own HR GPT - private, secure, and grounded in your knowledge.
A dedicated LLM assistant trained on your company's policies, JDs and playbooks. It drafts job descriptions and interview questions, screens and summarises resumes, and answers HR policy questions for employees and recruiters - inside a private, secure boundary where your data is never used to train external models.
Private by design
The power of a large language model, without giving away your data
Public chatbots are a compliance risk: paste a policy or a resume into one and you've potentially leaked confidential data and lost control of where it goes. HR GPT gives your team the same conversational speed inside a walled garden - grounded on your documents, governed by your permissions, and audited end to end.
Because it is grounded on your approved knowledge base with retrieval-augmented generation, every answer is drawn from your own policies, handbooks and JDs - not a model's guesswork - and comes with citations back to the source. Sensitive actions keep a human in the loop, so HR GPT accelerates your team without ever taking a decision out of their hands.
- Never trains external models - your policies and candidates stay yours, full stop.
- Grounded & cited - answers reference your source documents to reduce hallucination.
- RBAC & PII redaction - people see only what their role allows; sensitive data is masked.
- Human-in-the-loop - it drafts and suggests; your team reviews, edits and decides.
Private tenancy
Your assistant, your data boundary - isolated, encrypted and access-controlled.
Grounded answers
Every response cites the exact policy or JD it came from - verifiable, not vague.
Hours back per week
Drafting, screening and policy lookups that took hours now take minutes.
Audit trail
Every prompt, source and answer logged for review, compliance and tuning.
Key features
One assistant, across the HR and recruiting workflow
Six capabilities that put your knowledge base to work - for recruiters, HR business partners and employees alike.
JD & job-ad drafting
Generate role-accurate, inclusive job descriptions and job ads in your tone and format in seconds - refined from your existing library and ready for a recruiter to polish.
Resume screening & candidate summaries
Summarise long resumes into structured highlights, match them against the JD, and surface strengths, gaps and questions - so recruiters shortlist faster with a human making the call.
HR helpdesk & employee policy Q&A
Give employees instant, accurate answers on leave, benefits, expenses and policy - grounded in your handbook, deflecting routine tickets from your HR team around the clock.
Interview kit & question generation
Produce role-specific interview guides, competency questions and scorecards calibrated to the JD and seniority - consistent, fair and ready to hand to your panel.
Guardrails, citations & audit trail
Configurable guardrails keep the assistant on-policy, citations make answers verifiable, and a complete audit log records every interaction for compliance and continuous improvement.
How it works
From your documents to a trusted assistant in five steps
A secure, privacy-first build that keeps your data inside your boundary at every stage.
1 · Ingest & secure your docs
We securely ingest your policies, handbooks, JDs and playbooks into a private, encrypted knowledge store - with PII handling and access rules defined from the outset.
2 · Ground the model
We connect an LLM to that knowledge base using retrieval-augmented generation, so answers are drawn from your approved content - never used to train external or shared models.
3 · Configure roles & guardrails
We set role-based permissions, PII redaction, topic guardrails and citation requirements so each user sees only what they should and every answer stays on-policy.
4 · Integrate
We surface HR GPT where your people already work - Slack, Microsoft Teams, your ATS and HRMS - via native connectors, APIs and SSO.
5 · Deploy & monitor
We launch with training, then monitor accuracy, usage and feedback through the audit trail - tuning grounding and guardrails so quality improves over time.
Privacy & security
Built so your legal and security teams can say yes
HR data is the most sensitive an organisation holds. HR GPT is architected privacy-first - the reason it clears enterprise security review where public chatbots never could.
- No training on your data - your documents and conversations never leave your boundary or feed external models.
- Private tenancy & encryption - isolated instance, TLS in transit and AES-256 at rest.
- RBAC & PII redaction - least-privilege access and automatic masking of sensitive fields.
- DPDP, GDPR & SOC 2-style controls - consent, retention, DSAR handling and governance built in.
- On-prem & data residency - deploy in your VPC or on-premises with open-weight models where sovereignty demands it.
Meets your people where they work
Surfaced inside your existing tools through secure connectors and SSO.
Slack
Microsoft Teams
Your ATS
HRMS - SuccessFactors, Workday, Darwinbox, Keka
Email
SSO / SAML
REST APIs & webhooks
Web & internal portals
Pairs well with
Part of a connected, AI-native HR stack
HR GPT is stronger alongside the rest of the Jobtrix stack - and the recruiting teams behind it.
AI Solutions for HR
Bespoke intelligence layers - parsing, ranking, interview intelligence and predictive models.
Custom ATS
Embed HR GPT inside a bespoke applicant tracking system built to your exact workflow.
Digital Twin Candidates
AI candidate profiles and simulations for role benchmarking and interview practice.
Leadership Hiring
CXO, VP and board search, supported by AI-drafted briefs and interview kits.
RPO
Embedded recruiting teams that run your pipeline with HR GPT in their toolkit.
All Technology
See the full Jobtrix AI-native HR stack, engineered by recruiters.
FAQ
Personal HR GPT, answered
No. Your documents, candidates and conversations are never used to train external or shared foundation models. HR GPT runs in private tenancy: your knowledge base is used only to ground answers for your organisation, data is encrypted in transit and at rest, and prompts and outputs stay inside your boundary. We align to India's DPDP Act 2023, GDPR and SOC 2-style controls, and support data-residency requirements.
HR GPT is grounded on your approved documents using retrieval-augmented generation, so answers are drawn from your policies, handbooks and JDs rather than the model's memory. Responses carry citations back to the source paragraph, the assistant is configured to say when it doesn't know instead of guessing, and sensitive actions keep a human in the loop. Every interaction is logged for review and continuous tuning.
We are model-flexible - we select from leading commercial and open-weight LLMs based on your accuracy, cost, latency and sovereignty requirements. Deployment can be in a private cloud tenant or, where you need full control, in your own VPC or on-premises with open-weight models. Your grounding data always stays within your chosen boundary.
Access is governed by role-based permissions and SSO. Recruiters, HR business partners, employees and leadership each see only what their role allows - an employee asking a policy question can't reach confidential candidate data, and PII is redacted where it isn't needed. Every query is attributable through the audit trail, so access is both controlled and accountable.
Yes. HR GPT understands and responds in multiple languages, including major Indian languages and the languages of your global workforce, so employees can ask policy questions and recruiters can draft content in the language they work in - while grounding remains anchored to your approved source documents.
Give your HR team a GPT they can trust.
Private, grounded and cited - trained on your knowledge, governed by your rules, and never used to train someone else's model.