The problem

A professional identity usually lives in disconnected resumes, job portals, chat threads, and stale profile links. Every application asks for the same facts again, while tailored resumes risk drifting away from what is actually true. KnownIn makes one structured profile the source of truth and turns every downstream surface into a controlled projection of it.

Architecture

Interactive system view

See the product from purpose to implementation.

Follow the work, then switch views when you want the practical answer: who it is for and what it is built with.

Flow pausedStep 01 / 06

Step 01 of 06

Profile Truth is stored as…

Profile Truth is stored as structured, tenant-scoped data: profile fields, verified skills, experience, projects, education, certifications, and AI-named custom sections. Firestore is accessed through server-only data modules, with audit events recording every chat-driven change.

Full system map

The grouped dependency view stays visible so the infrastructure is readable at a glance.

System map

How the pieces connect

← read left to right

Diagram · loads on scroll

KnownIn is organized around one rule: Profile Truth is the source, everything else is a controlled projection. Chat and MCP clients can update the structured profile through authenticated server boundaries; the resume builder selects only verified evidence; applications keep private answers separate; and the public card exposes a sanitized view. Firestore holds the tenant data and audit trail, Storage holds rendered PDFs, Gemini parses updates and performs fit analysis, and Razorpay webhooks—not the browser—decide subscription state.

Outcome

38 MCP tools give connected AI clients a complete, auditable interface to the professional identity.
Live at knownin.com with public cards, chat-driven profile updates, application tracking, and resume exports.
Three ATS-focused PDF templates plus job-specific tailoring, fit analysis, and learning-gap reporting.
Profile, resume, application, and billing boundaries are enforced in server-side code rather than left to prompts or UI conventions.

Lessons learned

The profile should be the database of record; resumes, public cards, and applications should be derived views with stricter boundaries.
An MCP integration is a product surface, not a thin adapter. Authentication, scopes, CORS, tool documentation, and safe failure modes all matter.
Truthful automation needs explicit verification states. A skill that is merely mentioned cannot silently become resume evidence.
Private job-search context must be modeled separately from shareable identity data, even when both are useful to an AI assistant.