Meet Agent Skye: what AI in HR actually looks like
Most "AI in HR" is a chatbot bolted onto an old HRMS. Agent Skye is different: a governed AI agent that reads your skill graph, drafts a hire-versus-upskill decision in rupees, and does it all without ever seeing an employee's name or salary.
Ask ten HR vendors about their AI in 2026 and nine will point you at a chat box. Type a question, get a paragraph back. It looks like progress, but it moves nothing. The human still has to open five screens, pull the numbers, and make the call.
The shift underway is from AI that answers to AI that acts, agentic AI. An agent can take a goal ("this team is short a data engineer"), plan the steps, gather the evidence, and hand you a finished, costed recommendation. Inside SKILWI, that agent has a name: Skye.
What is AI in HR, really?
AI in HR is the use of machine learning inside HR software to draft, cost and recommend people decisions: screening candidates, detecting skill gaps, and choosing whether to hire externally or upskill internally. There are three generations of it, and most Indian HR tools are stuck on the first two:
| Generation | What it does | What it can't do |
|---|---|---|
| Rules & scoring | Ranks CVs by keyword match, flags leave anomalies | No reasoning; brittle to wording |
| Chatbot / copilot | Answers questions, summarises a policy | You still do all the work of deciding |
| Agentic AI | Plans a task, gathers evidence, drafts a costed decision for approval | Needs strong data governance to be safe |
Skye sits in the third row. It doesn't wait to be asked a question. It watches for the decisions your managers actually face and assembles the answer before they ask.
What Agent Skye does across the hire-to-retire arc
Skye is not a single feature; it is the reasoning layer that runs across SKILWI's modules. A few of the jobs it does:
- Reads the skill graph. Skye works on top of SKILWI's India-native graph, 28,000+ skills mapped to NCO 2015, ESCO v1.2 and the NSQF, so its recommendations are grounded in structured capability data, not vibes.
- Drafts the hire-vs-upskill call. When a gap appears, Skye computes the cost of hiring against the cost of upskilling someone you already have, and puts a rupee figure on each option.
- Suggests internal moves. Before you post an external role, Skye surfaces the employees already close to the target role and scores the readiness gap.
- Explains itself. Every draft comes with the reasoning and the numbers behind it, so a manager can see why, not just what.
The part that matters: Skye never sees your people's data
This is where most "AI in HR" quietly falls down. To answer a question, many tools ship employee records: names, salaries, sometimes PAN or Aadhaar, off to a third-party language model. Under India's DPDP Act 2023, that is exactly the kind of transfer you do not want happening silently.
Zero-Data Architecture. Skye reasons over opaque internal IDs and skill codes, never personal data. The language model sees "employee 7F3A needs skill NCO-2512.03", not "Priya Sharma, ₹18 LPA". Names, salaries and identifiers stay AES-256 encrypted inside SKILWI and are re-attached only after the model has finished, on your side of the wall.
The result: you get agentic AI on your workforce without your workforce's personal data ever leaving your control or entering a model's training path.
Human-in-the-loop by design
Skye drafts. People decide. Every recommendation Skye produces is a proposal that a manager reviews and then approves or rejects, accountability stays with a person, which is both good governance and, increasingly, a regulatory expectation. The agent removes the assembly work; it does not remove the human.
That single design choice is what separates a useful AI HR agent from a liability. An agent that acts unsupervised on people decisions is a lawsuit waiting to happen. An agent that does the legwork and hands a clean, explained draft to a human is leverage.
Is AI in HR worth it for a mid-market Indian company?
For a company of 200 to 2,000 employees, the maths is simple. Your HR team is small and your decisions are expensive, a wrong senior hire costs ₹30–40 lakh a year. Skye's job is to make sure the expensive decisions are made with evidence, in rupees, before you commit. You are not paying for a chatbot; you are paying for the decision-assembly a team twice your size would otherwise need.
Agent Skye is included in SKILWI's Develop and Decide tiers, from ₹500 per employee per month, so the agent that drafts your workforce decisions ships with the platform, not as a separate "AI add-on".
Common questions about AI in HR
- Will AI replace HR?
- No. It replaces the spreadsheet work between a question and a decision. The decision, and the accountability, stays with your people.
- Does the AI train on our employee data?
- No. Skye's Zero-Data Architecture keeps personal data out of the model entirely; it reasons over opaque IDs and skill codes.
- Is agentic AI safe under DPDP 2023?
- It can be, if it's built for it. SKILWI is DPDP-native: encryption at rest, consent, audit and erasure flows, and Skye inherits all of it.