AeroMind

The constraint

What the system had to solve before a line of it was designed.

Aviation organizations issuing pilot licenses, ATC clearances, and cabin-crew approvals needed a reliable, tamper-proof way to conduct psychometric evaluations. Their existing process was paper-based — tests were printed, invigilated manually, hand-scored, and filed in physical folders. Results took days to process, cross-referencing historical assessments against current scores was done by hand, and there was no way to share verifiable results with regulatory bodies or airline operators without physically dispatching documents. Beyond the operational drag, the fragmentation created real risk: a missed evaluation, a miscalculated score, or a misfiled record could result in an unqualified candidate passing an assessment that their actual profile should have flagged. The organization also ran periodic re-assessments for recurrent certification — with no system tracking evaluation history, there was no reliable way to detect drift in a candidate's cognitive or psychological profile over time.

Architecture

The stack read top down, from what the user touches to what keeps it running in production.

Modules delivered

Each one shipped production-grade, not as a demo surface.

• 01Test Builder — 12 module types covering cognitive, personality, aptitude, SJT and stress assessments
• 02Test Builder — adaptive difficulty, randomized ordering, timing controls and role-specific scoring weights
• 03Candidate Management — full profile lifecycle, document management and assessment scheduling
• 04Candidate Management — cohort and batch grouping with continuous evaluation history and drift detection
• 05Secure Examination — browser-locked, distraction-minimized test-taking environment with auto-save
• 06Secure Examination — real-time proctoring hooks, tamper detection and SHA-256 response signing
• 07Reporting — individual scored breakdowns with norm-group comparisons and role benchmarks
• 08Reporting — batch analytics, recurrent-assessment trend charts and disqualification audit trails
• 09Regulatory Output — DGCA and ICAO formatted PDF reports for submission to regulatory authorities
• 10Platform Foundation — RBAC across assessor, medical officer and admin roles, encrypted result storage

Scope delivered

What the build covered, end to end.

1. 016+ role profiles
2. 028+ report formats
3. 03DGCA/ICAO regulatory output

Design decisions

What we learned before building, and what the built system moved.

We began with a discovery phase alongside the client's chief assessor, psychometric officers, and the IT head — mapping the current paper workflow step by step, from test printing through manual scoring to physical filing. That map surfaced three critical requirements no off-the-shelf tool met: role-specific norm groups calibrated to Indian aviation workforce data, tamper-evident result storage that could satisfy a regulatory audit, and a test-taking environment that could be administered at distributed training centers without requiring local software installation. From these requirements we designed the data model first — candidate profiles, assessment records, response payloads, and result summaries as separate, immutable entities linked by foreign keys, so no re-assessment could overwrite a prior record. We chose PostgreSQL over a document store specifically for the relational integrity guarantees needed for regulatory audit trails. We designed the examination interface around DGCA examination guidelines, validating screen layouts and timing mechanisms with the client's psychometric officers before a line of implementation code was written. Every report format was prototyped as a PDF mock-up, reviewed against ICAO documentation standards, and signed off before the reporting module was built.

AeroMind replaced a fully paper-based psychometric evaluation process with an online platform covering the complete assessment lifecycle — from test design to regulatory submission. Assessment turnaround time dropped from multiple days to same-day, with automated scoring eliminating manual calculation entirely. The platform introduced continuous candidate history for the first time, enabling assessors to detect drift in cognitive or psychological profiles across recurrent evaluations rather than treating each assessment in isolation. Tamper-evident, SHA-256-signed result records satisfy regulatory audit requirements without requiring physical document dispatch. Role-specific norm groups and weighted scoring models, calibrated to the client's candidate population, gave the assessment results statistical validity that paper-based scoring lacked. The regulatory report generator produces DGCA and ICAO-formatted PDFs directly from stored results, removing a multi-day manual compilation step from the certification workflow.