Seven AI capabilities woven into HR, Payroll, Compliance, and Recruitment. None of them make the decision. They flag, score, parse, and prepare — so when you hit Approve, you're approving from a better starting point.
Five live today. One Premium Pack-bundled. One on the roadmap. Each maps to a specific module — AI in hrPLANR isn't a tab, it's where the work gets done.
Upload a resume (PDF, DOCX, scanned) and AI extracts name, contact, work history, education, skills, location, expected CTC, notice period. 92% field-level accuracy on Indian-format CVs. You verify before saving.
Before you close the cycle, AI scans every line for variances over 30% vs the 3-month rolling baseline. Department-level, role-level, employee-level. Flags surface in the cycle, AI suggests cause, you investigate.
Upload a PAN card, AI extracts the number and verifies against TRACES. Upload an Aadhaar, AI stores the last-4 in UI and encrypts the rest. Upload an offer letter, AI parses role, joining date, and salary. 30 min of data entry becomes 3 min of verification.
Before any statutory filing goes out, AI cross-checks the numbers against the source payroll, the prior cycle, and the rules. "PAN missing · 2 employees · TDS at 20% applies · Review" is a real flag — not a hypothetical. Available wherever filings happen.
A natural-language interface — in Slack, Teams, and the web app — to query your HR data. "How many people are on PL next week?" "What's our attrition rate this quarter?" "Who's about to complete probation?" — answered instantly, with the source data linked.
Signals across attendance patterns, performance, tenure, and engagement combined into an attrition risk score per employee. The conversation it enables — "I'm worried we're losing Anita in Q3" — happens before the resignation letter, not after.
Hiring funnel conversion rates. Cost-per-hire by source. Headcount projections by department. Time-to-productivity by role. The dashboards your CFO and CHRO actually look at — generated from real platform data, no Excel work.
Beyond extraction: verifying that uploaded documents are genuine. Aadhaar number checks, PAN-name matching, employment certificate validation against issuing entities. The compliance posture every IT lead asks about, automated.
Three concrete examples — across Payroll, Recruitment, and Compliance — that show what "human-in-loop" actually looks like in the day-to-day. AI prepares the work. You do the deciding.
Where does the inference happen, is your data used to train models, can the whole thing be turned off, and what about consent. All answered specifically below.
For Enterprise customers, all AI inference happens in-region on AWS Mumbai (ap-south-1). Employee data is not sent to overseas endpoints for processing. This matters for DPDP Act 2023 compliance and for sectors with data-residency contractual requirements.
Ours, theirs, or anyone's. We use the AI to do work *on* your data — we don't use your data to make the AI smarter. This is in our DPA, not just in a marketing claim. Your data leaves the inference call; no copies, no derivatives, no fine-tuning datasets.
Don't want AI in payroll? Turn it off. Want resume parsing but not fit scoring? Configurable. AI is opt-in at the feature level in tenant Settings — the rest of the platform works the same way whether AI is on or off. Nothing is locked behind it that shouldn't be.
Each employee sees which AI features apply to them — resume parsing during recruitment, attrition modeling during employment. Consent is captured per the DPDP Act's "purpose-limited" requirement. If they withdraw consent, AI processing stops for them.
We'd rather tell you the boundary today than hide it until the sales call. Three items committed for 2026. Anything beyond is exploration — we'll publish when we commit.
30-minute walkthrough — payroll anomaly detection, resume parsing, and the HR Assistant — with your data.