Data Collection & AI Transparency Disclosure

Explicit declaration of what data TryCareerLens collects, how AI models process it, and data retention windows.

Effective Date: August 4, 2026 | Last Updated: September 15, 2026

Transparency First

At TryCareerLens, we believe software engineers should know exactly what data is collected, how it is analyzed by Artificial Intelligence models, and who can view it. This document provides a line-by-line breakdown of our data collection practices.

Complete Data Collection Inventory

Data Category	Specific Fields Collected	Collection Trigger	Primary Purpose
Identity Data	Full Name, Email Address, Phone Number, City, State.	Account Registration	User authentication, candidate profile creation.
Resume Data	PDF/Word Document file, extracted skills, work history, degree tier.	Resume Upload	Parsing claimed technical trajectory & baseline background.
GitHub Metadata	Public repo names, language bytes, commit frequencies, Docker/k8s configs.	GitHub Connect (OAuth)	Analyzing real-world codebase output vs claimed skills.
Assessment Telemetry	Scenario responses, quiz scores, timing flags, integrity events.	Validation Tests	Evaluating technical comprehension & claimed vs validated skills.
Payment Telemetry	Razorpay order ID, payment ID, transaction status, billing amount.	Razorpay Checkout	Provisioning digital credits & recruiter plan subscriptions.
Placement Outcome Data	Anonymized time-to-hire, engineering score correlation, offer acceptance metrics.	Recruiter Offer Response & Placement Confirmation	Aggregated analytics to train assessment quality, optimize matching algorithms, and elevate evaluation accuracy.

AI Processing & Model Subprocessors

TryCareerLens utilizes Large Language Model (LLM) providers to parse documents and generate candidate feedback:

Anthropic PBC: Resume parse, assessments, mentor chat, autopsy, live interviewer, batch evaluation.
Voyage AI: Skill embeddings for matching (identity fields stripped).
Sarvam AI / ElevenLabs: Voice interview speech-to-text and text-to-speech.
Groq / OpenRouter: Optional alternate LLM routing when configured.
Razorpay, Cloudflare R2, Railway, GitHub, Resend, Google Analytics: Payments, private file storage, hosting, OAuth, transactional email, optional analytics (consent-gated).

Strict Non-Training Guarantee: Your uploaded resume content, GitHub code snippets, and assessment responses are processed via API endpoints under enterprise zero-data-retention terms. Your data is NEVER used to train public AI models.

Data Retention & Account Erasure Policies

Active Account Retention: Candidate profiles and assessment history are retained as long as your account remains active.
Right to Complete Erasure & Account Closure: Account closure from `/account` deletes Postgres rows, Redis session/OTP/GitHub keys, R2 resumes/avatars/voice interview audio, and embeddings. Optional purposes can be withdrawn without full erasure from Account → Privacy.