TalentIQ reads the public record for the moments that change that — a funding round that goes quiet, a layoff filing, a Naukri profile refreshed at 11pm, commits that stop landing — and tells your recruiters who to call first, and why.
Built for Indian hiring teams. Data stays in ap-south-1. Every profile carries the sources it was built from.
A recruiter's week spent sourcing one role, per LinkedIn's Global Talent Trends.
Share of Indian white-collar hiring that is non-tech — where GitHub tells you nothing.
Of Naukri offers go to candidates who refreshed a profile in the prior 30 days.
Of double-contacted candidates lose interest in the employer. Deduplication is a product feature.
What it watches
Every sourcing tool in this market was built for engineers, because engineers leave a public trail of work. Two thirds of Indian hiring is not engineers. TalentIQ runs a second signal set for sales, finance, operations, HR and marketing, where the evidence is inferred rather than committed.
How the ranking is built
Every candidate carries a 0–100 score against one specific role, made of five parts that are always shown separately. A recruiter who disagrees can see which component is wrong, and say so — that disagreement is the training data.
Non-tech roles run a different split — skills drop to 25 and company experience rises to 25, because where someone has worked predicts more than a declared skill list when there is no code to read. TA leads can move the weights per role family, and every change is versioned.
Until a workspace has logged 500 hiring outcomes the weights stay rule-based and human-set. After that a gradient-boosted model takes over, retrained quarterly, and only shipped if it clears 75% precision on the top 20 candidates per role family. Below that, the previous model stays.
Strong match. Queue for outreach.
Good match. Read the profile first.
Partial. Pipeline material, not priority.
Weak. Shown only when the pool is thin.
Scores are tested every six months for drift by gender, age band and institution. A failing dimension blocks the next training run.
The part nobody demos
Sending is the easy half. Staying deliverable is the hard half, and it is where automated sourcing tools quietly destroy their customers’ LinkedIn accounts and sending domains. TalentIQ ships with the ceilings already set below the platform limits.
| Control | Platform limit | What TalentIQ does |
|---|---|---|
| InMail per recruiter, per day | 20 (Recruiter Lite) | Bot sends up to 16, leaving the remainder for the recruiter’s own messages. |
| Connection requests per week | 100 | Capped at 70 per account, rotated across seats so no single account spikes. |
| Email bounce rate | 3% | Outreach to that segment pauses automatically and the list is flagged for cleaning. |
| Spam complaint rate | 0.1% | Immediate stop on the sending domain, before Gmail makes the decision for you. |
| New sending domain | 4 weeks | Warmed from 20 to 500 a day on a subdomain, never the company’s primary domain. |
| Repeat contact | 21 days | The bot checks the contact log first. A second send inside the window is blocked, not queued. |
| Candidate replies | Instant | Automation stops on that candidate and the recruiter takes over. No bot ever answers a human twice. |
| Claim on a candidate | 60 days | One recruiter owns the relationship. Others can read the profile but cannot message without a transfer. |
Weekend sends are off by default. Every message, reply and suppression is written to the candidate’s timeline, so a TA lead can reconstruct exactly what a person was told and by whom.
After the offer
A signed offer in this market is a three-month option the candidate holds over you. Notice periods of 90 days are standard in IT, and offers made to those candidates drop out around a third more often than offers to people on 30 days. TalentIQ scores that exposure from the day the letter goes out.
| What drives a drop | Weight | What TalentIQ reads |
|---|---|---|
| Window exposure | 25 | Days still to run against the notice length. The single largest predictor, because everything else needs time to happen. |
| Engagement | 25 | Reply latency, opens, documents submitted, background checks moving. A candidate going quiet is the earliest reliable tell. |
| Offer strength | 20 | Where the offer sits in the market band, and whether a notice buyout was included. |
| Counter-offer exposure | 15 | A healthy, well-funded employer is more likely to buy someone back. This is the exact inverse of the sourcing signal. |
| Anchors | 15 | Met the team, has a referrer inside, signed rather than a verbal yes. |
The schedule is generated from the notice window and each touchpoint has an owner, because the thing that saves an offer is rarely another email from the recruiter.
Hiring manager calls. The first 90 days, and what the role actually owns.
Resignation confirmed and the last working day put in writing.
Introduced to two people on the team, one of them a peer rather than a manager.
Check-in on the notice period and any pushback from the employer. This is when buybacks land.
Week-one plan and a buddy. By now they should be able to picture the first Monday.
The moment a candidate accepts, third-party intent crawling for that person is switched off. Every signal after that point comes from how they deal with us.
This is a deliberate limit, not an oversight. Monitoring an accepted candidate’s job-seeking is surveillance of someone who has already entered a relationship with you, and it changes the purpose their data was collected for. It would also be the fastest way to lose them if they ever found out.
| Where the market sits | Rate |
|---|---|
| Typical Indian joining ratio | 65–75% |
| A healthy one | 85–90% |
| Offer acceptance, Ashby benchmark | 78% |
| Extra dropout on a 90-day notice | +35% |
Every drop is coded against a fixed taxonomy — counter-offer, competing offer, compensation, role scope, relocation, timeline, personal. “Candidate declined” is not data. The reason decides which part of the sourcing model moves, which is how post-offer outcomes end up improving who gets surfaced in the first place.
India first
TalentIQ processes data about people who never signed up for it. That is workable under legitimate use, and only workable if the notice, the erasure path and the retention clock are real product surfaces rather than a policy page.
Every first contact carries one line saying where the profile came from and how to be removed. Replying REMOVE works, and works in ten minutes.
A full notice goes out before any further data is collected — what is held, why, and for how long.
Deleted from every pipeline and suppressed from future crawls. Permanent unless the person opts back in themselves.
A candidate who went cold gets an automated re-consent mail. No answer in 14 days and the record is deleted.
Hired candidates are retained as an employment record, then reduced to anonymised outcome data for model training.
| Where the law actually stands | Date |
|---|---|
| DPDP Act passed | Aug 2023 |
| DPDP Rules notified, Data Protection Board established | Nov 2025 |
| Consent Manager framework operational | Nov 2026 |
| Full compliance required | May 2027 |
The substantive obligations — notice, consent, erasure, breach reporting — are phased in and become enforceable in May 2027. That is a build window, not an excuse: any workspace that starts logging consent now arrives at the deadline with an audit trail instead of a migration.
Indian candidate data sits in ap-south-1 by default, with a dedicated tenant available so no two customers share a candidate store. No third-party enrichment call is made without a processing agreement behind it.
Who it is for
The same candidate looks different depending on who is asking. TalentIQ shows each of them only what they can act on.
Ranked candidates with the reasoning attached, and one keystroke to shortlist, decline with a reason, or hand to the bot.
Pipeline health, stage conversion, stale alerts at 45 days, and a compliance report that downloads without a ticket.
A shortlist above the threshold, not a dump. Approve to contact, or send it back with a reason.
Outbound hiring without an agency retainer, and a straight answer on what the pipeline will produce this month.
Point TalentIQ at a requirement you are struggling to fill, and compare its first twenty against whatever your team found last week.