How is BeeTrustScore different from AMCAT or eLitmus?
AMCAT, eLitmus and SHL publish single-test percentiles — one number, one
moment, opaque. BeeTrustScore is a composite built from oral, domain-specific,
case-study, role-play and video assessments, peer reviews and feedback over time, and
every point is decomposable via SHAP. Colleges can defend the readiness picture,
students can dispute the components, and the score grows with the track record — not
from a single Saturday.
Is the AI deciding my future?
No. The AI grades fluency on a published rubric, checks domain work in
sandboxes where applicable, and verifies liveness on video. Every claim is grounded in
a specific signal. A BeeTrustScore-licensed reviewer signs off on disputes. Students
and colleges see the score plus the decomposition — humans make the calls, not the
model.
How is my data protected?
Per-purpose granular consent at intake — biometric consent stays separate
from generic consent, per DPDP Section 5. Videos are encrypted, retention is 12 months by
default, deletion is one click. We don't sell data, we don't train third-party models on
your submissions, and APAAR / ABC / DigiLocker stay your wallets — we read, we never
replace.
Is the score biased?
We publish an annual bias audit — caste, gender, region, accent,
disability. Delta thresholds are stated upfront. If a dimension drifts, we freeze model
updates until the cause is found. Bias monitoring is built into the platform-admin
surface, not an afterthought.
What if a student disputes a score?
Every score component is challengeable. Students can flag a specific
dimension (e.g. "my coding-72 should be higher — see this commit"), a licensed reviewer
adjudicates within 48 hours, and the audit trail is preserved. Disputes feed into the
model-drift dashboard so we can catch systemic issues early.
Who is BeeTrustScore for, day one?
Pilot scope is colleges and their students at NAAC-A or
equivalent institutions — the college placement and academic teams running the cohort,
and the students building their score across three years. Phase 1 is one cohort
end-to-end — anchor, assess, verify, score, improve. Phase 2 opens employer-facing
rails. Phase 3 widens to white-label trust-score APIs.
How does this plug into APAAR / ABC / DigiLocker?
APAAR is the anchor — every BeeTrustScore profile is bound to a verified
APAAR ID via DigiLocker. ABC credits feed into the dimension breakdown (course
performance, verified). We read from the wallets; we never replace them. If a student
leaves the platform, the underlying identity stays with the government rails — as it
should.
Is BeeTrustScore Kerala-only?
Phase 1 pilot is Kerala-first — we build small, test, learn with a
focused set of Kerala colleges before expanding. India rollout follows Kerala validation;
global expansion follows India traction. The platform itself isn't geographically locked —
but our acquisition and pilot focus stays Kerala while Phase 1 ships.
How does BeeTrustScore relate to KSUM / IEDC?
We're in pilot conversations with Kerala's startup ecosystem partners —
including the Kerala Startup Mission (KSUM) and its 453+ IEDC college network. No formal
partnership has been announced. We share KSUM's goal of measurable student outcomes across
the IEDC ecosystem, and our Phase-1 pilot scope is designed to plug into that network. Any
partnership shape will be announced once it's concrete on both sides.
What about students who aren't 18 yet?
Phase 1 is 18+ only — we honor DPDP Act Section 9, which requires
parental consent for processing minors' data. The parental-consent surface ships in Phase 2
once we have a verified DPDP §9 flow (parent identity verification + scoped consent +
per-purpose granularity). Until then, the platform onboards students at or above 18 only.
