If you hire in India and you are not on Greenhouse or Workday — and most of the Indian mid-market is not — you are probably on Zoho Recruit or Keka. Between them, the two systems cover a very large share of Indian IT services, product startups, BPO, and SME hiring.

Both are solid ATSes with active APIs. Neither has the same integration surface as Greenhouse. That changes the integration playbook in specific ways this article unpacks.

The right metric for an India integration is not "feature parity with the US ATS." It is "does the recruiter ever need to leave the system to act on the AI scorecard?" If not, the integration is working.

Why Zoho Recruit and Keka dominate Indian mid-market hiring

Two factors. First, pricing — both are dramatically cheaper than US-headquartered competitors at the seat counts Indian mid-market companies need. Second, Indian payroll and compliance integration — Keka in particular is a full HRIS that handles statutory compliance (PF, ESI, professional tax, gratuity) out of the box, which a non-Indian ATS does not.

That means the integration design must respect a constraint: recruiters are not going to leave Zoho or Keka to use a separate AI screening tool. The tool has to come to them.

Zoho Recruit — the integration surface

Zoho Recruit exposes a REST API that covers candidates, jobs, applications (called "candidate-job associations" in Zoho's vocabulary), and custom modules. The API supports OAuth 2.0 with scoped tokens.

The objects that matter for an AI interview integration:

Candidate — the person and their resume.
Job Opening — the role.
Candidate-Job Association — the equivalent of a Greenhouse application.
Interview — Zoho's first-class interview object, which can be scheduled, completed, and have feedback attached.
Custom Module — Zoho allows custom modules and fields, which is how AI-specific metadata gets stored cleanly.

Zoho Recruit's webhook coverage is more limited than Greenhouse's. The integration should expect to do more polling and less event-driven triggering.

The Zoho Recruit integration pattern that works

Define a custom module — call it "AI Interview" — with fields for status, invite timestamp, completion timestamp, transcript URL, and overall recommendation.
Define a workflow rule in Zoho that, when a candidate is moved to a specific stage (commonly "AI Screen"), creates an AI Interview record linked to the candidate-job association.
The integration polls Zoho for new AI Interview records every few minutes and triggers the interview invite.
On interview completion, the integration writes the scorecard fields back to the AI Interview record and also writes a structured Interview feedback record against the Zoho Interview object, so the scorecard appears in the recruiter's normal Zoho view.

The recruiter never needs to know an external system exists.

Keka — the integration surface

Keka is a fuller HRIS with hiring as one module. Its public API covers candidates, jobs, applications, and assessments. The API uses bearer-token authentication.

Keka's interview workflow is opinionated: each application moves through a configured pipeline of stages, and each stage can have feedback forms attached. The AI interview integration plugs into this stage system.

The objects to know:

Job — the role.
Candidate — the person.
Application — the candidate-job pairing.
Pipeline Stage — the workflow position of the application.
Feedback Form — the structured scorecard attached to a stage.

The Keka integration pattern that works

Keka's pattern is closer to Greenhouse than Zoho's is.

Add an "AI Screen" stage to the relevant job pipelines in Keka.
Define a feedback form for that stage with fields that mirror the AI scorecard rubric.
Configure a webhook on the stage transition to call the AI interview integration.
The integration receives the webhook, invites the candidate, runs the interview, and writes a feedback form submission against the application at the AI Screen stage.

Recruiters see the scorecard in the same Keka view they use for human-conducted interview feedback.

Hand-off fields that matter — the India-specific list

Most of the hand-off fields are the same as the Greenhouse list, but four India-specific items are worth calling out.

Notice period. Indian hiring is dominated by candidates serving 30, 60, or 90-day notice periods. The AI should be prompted to surface this early and confirm it; the scorecard should record it as a structured field.
Current and expected CTC. Indian compensation conversations are open and structured around fixed/variable splits and total CTC. The interview should capture both as structured fields for the recruiter's downstream offer process.
Location flexibility. Tier-1 hiring often involves candidates from other cities. Willingness to relocate, hybrid expectations, and existing location should be captured.
English fluency self-report and observed. For roles where client-facing English is required, the AI's observed assessment is usually more reliable than the candidate's self-report. Both should be recorded.

These are not optional in the Indian context. A scorecard that does not surface them forces the recruiter to do a manual call anyway, which negates the integration's value.

The bilingual interview question

A meaningful share of India hiring — particularly for sales, BPO, and field roles — happens partly in Hindi, Tamil, Telugu, Kannada, Bengali, or other regional languages. For technical and product roles, English is the default.

The integration design has to handle three modes:

English-only. Default for technical, product, and most leadership roles.
Mixed (code-switching). Common in many real conversations; the AI should not penalise candidates for it.
Regional-language primary. For field, BPO domestic, and some retail roles, the interview itself should be conducted in the candidate's primary language.

The recruiter-facing scorecard should be in English regardless, with the original-language transcript available for review.

A failure mode specific to India hiring

The single most common failure mode in Indian high-volume hiring integrations is candidate no-show without notice. Conversion rates from "interview invited" to "interview started" can be substantially lower than in GCC or US contexts.

The integration design should account for this with:

A short invite-to-start window (typically 48 to 72 hours).
One automated reminder, not three.
A clear expired state that triggers automatic stage movement in the ATS so the recruiter is not chasing ghosts.

Trying to engineer around the conversion rate with aggressive reminders typically reduces it further. Better to accept the rate, optimise the funnel for it, and let recruiters spend their time on candidates who actually engage.

Pricing reality

A working AI interview integration on Zoho Recruit or Keka should not double the per-hire cost of the ATS itself. The pricing model that works in the Indian mid-market is per-interview, not per-seat, billed in INR with GST handled correctly.

This is a procurement issue more than an engineering one, but it matters: an integration that delivers technically but breaks on pricing economics will not survive procurement.

A 30-day rollout for Zoho Recruit and Keka teams

Week 1. Decide on one role family. For Zoho, set up the custom module and workflow rule. For Keka, add the AI Screen stage and feedback form.
Week 2. Run 20 candidates through, with each scorecard manually reviewed by a senior recruiter alongside their independent notes.
Week 3. Calibrate the rubric where the scorecard and the recruiter disagree.
Week 4. Expand to all roles in that family across all recruiters.

The pattern is the same as on Greenhouse. The discipline of one-role-at-a-time rollout is what makes it stick.

Where to go next

To see what the candidate side of the interview looks like — including the bilingual handling that India hiring needs — the Voxxhire demo walks through it in under three minutes.

For the same playbook applied to Greenhouse, see our Greenhouse AI interview integration playbook. For a broader take on scaling first-round screening, see BPO hiring at scale.

For an example of an early-stage company running structured AI hiring at low volume but with India-style discipline, see the Exara AI design partner case study.

Zoho Recruit and Keka for India Hiring: An Integration Playbook