VerifAI
VERIFIED PLACEMENT INTELLIGENCE

Turn student evidence into explainable placement decisions.

VeriAI brings resumes, marksheets, GitHub, and coding-platform evidence into one structured profile, helping placement teams verify eligibility, understand readiness, and create transparent shortlists.

Single-college placement workspace · Verified transcripts & code evidence · Zero black-box ranking

https://verifai.app/tpo/candidates
JD: "Full Stack Engineer — React, Node.js, TypeScript, Min 7.5 CGPA, No Backlogs"
Drive #2026-FSE · 34 Verified Students Evaluated · Matched Criteria
Marksheet VerifiedExplainable Match
TPO Criteria: Are you filtering for unplaced students only or all eligible candidates?
AS
Aarav Sharma
CSE · CGPA 8.95
98%
Match
ReactNode.jsTypeScript
Marksheet OCR420 commits
PP
Priya Patel
IT · CGPA 8.42
94%
Match
ReactTypeScriptNext.js
Zero BacklogsTop 8% Rating
RV
Rohan Verma
AIML · CGPA 8.15
89%
Match
Node.jsPythonReact
Verified Resume310 commits

Designed for Campus Ecosystems

Two perspectives. One trusted evidence pipeline.

VeriAI connects student readiness with placement office decision-making through transparent, auditable profiles.

Self-Serve Portal

For Students

Understand your placement readiness, strengthen your evidence, and see practical next steps before the next opportunity.

  • Consolidate resumes, marksheets, GitHub, and LeetCode into one auditable profile.
  • Receive deterministic skill extraction and practical guidance on missing proficiencies.
  • Verify academic credentials once with encrypted, private document storage.
TPO Console

For Placement Teams

Verify academic eligibility, review coding evidence, and build explainable shortlists from a job description.

  • Enforce marksheet-verified CGPA and backlog rules over unverified self-claims.
  • Parse job descriptions instantly and rank candidates with clear matched vs. missing criteria.
  • Export audit-ready shortlists in CSV format for visiting recruiting teams.

Workflow Overview

How VeriAI Works

A structured four-stage pipeline from primary document submission to audit-proof candidate shortlists.

01

Collect evidence

Students submit resumes, marksheets, and coding-profile details.

Multi-Source Submission
02

Verify and analyze

VeriAI extracts skills, analyzes coding evidence, and uses marksheet data as the academic source of truth.

Academic Truth & OCR
03

Match to opportunities

TPOs add a job description and receive ranked candidates with transparent match evidence.

Explainable Ranking
04

Guide improvement

Students receive readiness signals and practical actions to improve.

Actionable Feedback

Verification Integrity

Not just resume keywords.

VeriAI replaces speculative resume scans with multi-source proof across academic transcripts, live codebases, and deterministic criteria.

Verified academics

Marksheet-derived CGPA takes priority over self-reported values.

Automated OCR extracts cumulative CGPA and active backlog count directly from university marksheets, rejecting manual inflation.

Coding evidence

GitHub and LeetCode data add practical technical context.

Inspects public repositories, commit patterns, and algorithmic problem solve distributions rather than taking claimed skills on faith.

Explainable matching

Every shortlist shows matched skills, missing skills, and eligibility evidence.

TPOs and recruiters get transparent reason codes for why each candidate is placed in rank order—no black-box scoring.

Private document handling

Student documents are stored privately with controlled access.

Academic records and resumes reside in private AWS S3 storage, accessible only via time-bounded HMAC signed tokens.

Infrastructure & Security

Built with a practical, secure cloud foundation.

Engineered with dedicated cloud primitives on Amazon Web Services to guarantee private document storage, isolation, and reliable placement execution.

Production Cloud Topology
AWS ap-south-1·Controlled IAM Scopes
01 Client

Next.js Frontend

App Router client serving student portals, TPO console, and real-time candidate search views with client-side token management.

HTTPS / TLS 1.3
02 Gateway

Amazon API Gateway

Single secure ingress point managing request routing, authentication header verification, and rate limiting.

Managed API Routing
03 Services

FastAPI on Amazon EC2

Dedicated microservices handling PDF OCR parsing, GitHub evidence analysis, and deterministic candidate ranking.

Python 3.12 / Asynchronous
04 Database

PostgreSQL Database

Relational schema storing structured candidate records, placement drives, historical batch results, and audit trails.

Encrypted at Rest
Private Amazon S3 Storage

Academic marksheets and student resumes reside in a strictly private AWS S3 bucket. Access is granted only via server-signed, time-bounded tokens with zero public bucket endpoints.

AWS IAM Controlled Access

Application roles follow the principle of least privilege. Backend execution nodes have restricted, role-bound permissions to read and write document prefixes with no shared root credentials.

Product Stage & Vision

Built for placement teams. Designed to scale across institutions.

VeriAI is currently demonstrated as a secure college placement workspace. Multi-college onboarding and tenant isolation are planned as the next platform layer.

Current Deployment

Dedicated College Instance

Operates as an isolated institutional workspace with dedicated database storage, campus-specific criteria rules, and separate student/TPO roles.

Next Architecture Layer

Multi-Tenant Institutional Scaling

Extending the core verification pipeline with tenant-isolated subdomains, institutional policy configs, and consortium-level recruiting portals.

FAQ

Frequently Asked Questions

Clear answers on VeriAI's architecture, evidence verification pipeline, and institutional deployment model.

Have more questions? Explore our Interactive Live Demo or access the TPO Console.

See an explainable shortlist in action.

Explore an isolated demo workspace with synthetic student profiles and role-based candidate matching.

Interactive demo workspace · Synthetic candidate datasets · No production changes