Proof
Evidence you can check, measured on your process
Most vendors prove a result with someone else's story. A number measured on another company's hiring or another company's paperwork tells you almost nothing about yours. So the proof here is the method, the mechanics and the controls, and then one bounded pilot on your own work.
- Baseline before pilot
- 8 scoring dimensions
- Same inputs, same score
- Data held in India
Last updated Reviewed by Hab Business Solutions
Short answer
What counts as proof at Hab?
A figure you measured on your own process, against a baseline recorded before anything was deployed, using the same definitions on both sides. Everything else on this page exists to make that measurement possible and honest.
The method
How a result gets proven
Five steps, agreed in writing before work starts, so the test has an edge and a flat result stays reportable.
Step 1
Baseline
Before anything is deployed, we measure the process as it runs today: hours spent, time to shortlist, how much of the applicant pool is genuinely read, and how often a shortlist survives the hiring manager.Step 2
Pilot
One bounded process, real roles, real applicants. The scope and the go-live date are agreed in writing before work starts, so the test has an edge.Step 3
Measure
The same figures captured at baseline are captured again during the pilot, on the same definitions. Shortlist acceptance is tracked alongside speed, because a faster shortlist that hiring managers reject is not an improvement.Step 4
Compare
Baseline against pilot, with the workings attached. Where the result is flat or worse, that is what the report says.Step 5
Decision
You decide whether to scale, adjust, or stop. The evidence for that decision is yours, measured on your process, not borrowed from a case study about someone else's.
The mechanics
What sits under a candidate score
Every rank is built from eight named dimensions, with the evidence attached to each one. Nothing about the score is a black box you have to take on trust.
Skills
Mandatory and optional skills weighted separately. Missing must-haves are surfaced, never used to silently drop a candidate.
Tools
Named platforms, frameworks, and systems, matched through aliases so equivalent terms count as the same requirement.
Experience
Depth and recency of relevant work. Employment gaps are recorded as neutral facts, not penalties.
Education
Degree-aware matching across supported degree variants and branches, rather than exact-string comparison.
Certifications
Role-relevant certifications identified and credited against the requirement they satisfy.
Projects
Delivered work that evidences a requirement even when the resume never uses the requirement's wording.
Title similarity
How closely prior titles map to the role, allowing for the fact that titles mean different things across companies.
Semantic similarity
Overall meaning-level fit between the resume and the job description, beyond keyword overlap.
The controls
Security and governance facts, stated as they are
Checkable before a pilot, not after.
In-region data residency
Customer data is stored and processed in India (Google Cloud asia-south1).
Keys stored as hashes
API keys are held as SHA-256 hashes, never as recoverable secrets.
Malware scanning by default
Every uploaded document is scanned before it enters the evaluation pipeline.
Time-bounded support access
Support access is explicit, read-only unless otherwise granted, and expires within 24 hours.
Tenant isolation
Multi-workspace collaboration with row-level security between tenants.
SOC 2-aligned controls
Controls are built to SOC 2 alignment, with independent certification on the roadmap. We say roadmap because it is a roadmap.
What we do not claim
The numbers we have not earned the right to publish
Hab methodology
No benchmark, no percentage
Hab Business Solutions does not publish a shortlist accuracy percentage or a speed multiple for MinMaxHR. We have not run an independent benchmark that would justify one, and a vendor-reported figure is not evidence. What we publish is the scoring method, the eight dimensions behind every score, the evidence under each dimension, and the measurement process a buyer can run on their own historical candidates.
Product capability
What is true of the product today
- MinMaxHR is a decision intelligence layer for recruitment. It sits next to an ATS, not in place of one.
- CandidRanker is the ranking and evaluation engine inside MinMaxHR.
- It is not an applicant tracking system, not a recruitment agency, and not an automatic rejection engine.
- Scoring is deterministic: the same job description and the same candidates produce the same order every run.
- No candidate is ever auto-rejected. Every decision is a recorded human action.
Questions
What buyers ask about proof
Does Hab publish customer case studies?
Not yet. Client work is under agreement, and a case study written without the client's sign-off is marketing, not evidence. What is published here is the measurement method, the scoring method behind MinMaxHR, and the security facts a buyer can check.
Why is there no accuracy percentage or speed multiple?
Hab Business Solutions does not publish a shortlist accuracy percentage or a speed multiple for MinMaxHR. We have not run an independent benchmark that would justify one, and a vendor-reported figure is not evidence. What we publish is the scoring method, the eight dimensions behind every score, the evidence under each dimension, and the measurement process a buyer can run on their own historical candidates.
How can a buyer verify a result before committing?
Run one bounded pilot on your own historical candidates or your own document backlog. Baseline first, then measure the same figures on the same definitions, then compare. The evidence belongs to you.
Where is hiring data stored?
Customer data is stored and processed in India, in Google Cloud asia-south1. Uploaded documents are malware scanned before they enter the evaluation pipeline, and support access is time bounded.
Measure it on your own process
One bounded pilot, your roles or your documents, your baseline. Hari Bharathi runs the first call.
No retainers to start · Pilot-first · Human-in-the-loop governance