Product Overview

What Kashif is

Kashif is an inbound hiring platform for small and mid-size companies. It replaces the spreadsheet-plus-email-plus-five-tools stack that most SMBs use to hire, with one workspace that takes a role from "we need to hire someone" to "they start Monday".

The product's promise, in one line:

The hiring platform that never ghosts and always shows its work.

Three commitments follow from that positioning, and the whole product is built around them:

1. Every applicant gets an answer. Confirmation on apply, a decision notification on every stage change, a private status page they can check any time.
2. Every AI decision can be shown to a lawyer. Scores come with cited evidence, a non-removable fairness rule set, optional blind screening, and a per-role compliance export.
3. A founder can go from blank page to booked interviews in one sitting — and pay for it self-serve, without talking to sales.

Who it's for

• Primary segment: small/medium companies hiring salaried roles (not agencies, not high-volume hourly).
• Primary domain: IT and tech roles first — screening defaults, calibration, and skill matching are tuned for technical candidates, though the platform is domain-agnostic.
• Primary users: founders, office managers, and small HR/recruiting teams who hire occasionally and don't want an enterprise ATS.

The core motion (inbound)

Post a job  →  Applicants apply on a branded careers page  →  Auto-screened + AI-ranked with
reasons  →  Team reviews the shortlist  →  Interviews are self-scheduled  →  Hire, with every
candidate kept informed at each step.

Outbound web-search sourcing (search + enrichment) is a supporting flow that feeds the same unified pipeline; it is not the front door.

What makes it different

• Explainable AI with a paper trail. Every fit score carries evidence citations and a full provenance record (provider, model, prompt version, input hash, output, cost). Most SMB ATSs give you a black-box number or nothing.
• Workspace-owned communication. Candidate email and interview invites go through the company's own connected Google Workspace or Microsoft 365 mailbox — better deliverability, real calendar invites, and the candidate hears from the company, not a platform relay.
• Defensible by default. A hardcoded fairness preamble the AI can never be told to ignore, deterministic knockout rules that keep hard gates under human control, optional blind screening, and an exportable compliance pack.
• SMB-honest pricing. Metering is on units a founder understands — active published jobs and applicants screened — not seats or sourcing credits.
• Cost-controlled AI. A Postgres-backed job queue with per-org caps, per-job/day screening limits, circuit breakers, and dead-lettering means AI spend can't run away.

Plans & pricing

Kashif has four plan tiers. Entitlements are metered on active published jobs and applicants screened per month (the units SMBs understand):

Plan	Active published jobs	Applicants screened / month	Typical fit
Free	1	25	Publishing your first role
Starter	3	150	A few concurrent roles
Pro	10	750	Active, growing hiring
Enterprise	Unlimited	Unlimited	High-volume / scaling teams

• Every new workspace starts on a 14-day full-feature Pro trial — no card required. At the end of the trial it moves to Free unless a plan is purchased.
• Billing runs through Moyasar (Saudi/MENA gateway: mada, Visa/Mastercard, Apple Pay, STC Pay). Prices are configured per deployment (in SAR by default); the in-app plan cards show indicative pricing for the deployment.
• Per-customer limit exceptions can be granted by platform operators without a code change.

See the Billing & plans guide for details.

Glossary

Term	Meaning
Workspace / organization	A tenant. Every record is scoped to one workspace; data never crosses workspaces.
Pipeline	The stages a candidate moves through for a job: applied → shortlisted → interview → offer → hired (or rejected).
Pipeline entry	One candidate's position in one job's pipeline.
Applicant vs. candidate	An applicant applied via the careers page; a candidate is the underlying person record (may also be sourced).
AI fit score	A 0–100 score with confidence, summary, strengths, gaps, risks, and cited evidence, produced against the job's calibration.
Calibration	The role's success criteria (must-have/nice-to-have skills, deal-breakers, target companies, rubric) that the AI scores against.
Knockout rule	A deterministic hard gate on an apply question (e.g. "not authorized to work → decline") evaluated before any AI call.
Screening	Auto AI evaluation of an inbound applicant.
Talent pool	Saved candidates kept for future roles.
Careers page	The public, branded, per-job apply page; the company careers page lists all of a workspace's open roles.
Compliance pack	A per-role export of every screening decision (scores, evidence, prompt versions, stage history, consent, audit trail).