Freddy Reyes logo
← All projects
JobsAI MCP · Private project

A clearer path from opportunity to next step.

JobsAI brings job discovery, evidence-based fit analysis, contacts, and follow-ups into one personal workspace. Semantic AI search finds relevant context; deterministic logic keeps core decisions grounded in explicit criteria.

Explore capabilities See how it works

JobsAI is a private project and is not publicly available.

JobsAI dashboard showing fictional opportunity counts, meetings, and follow-up activity.
One workspace for opportunities and next steps. Sample data is fictional.

The product

Keep the entire search connected.

Relevant roles are only the beginning. JobsAI connects each opportunity to résumé evidence, company and contact research, interviews, calendar events, and follow-up actions. Review queues, list and board views, daily planning, and reporting make the work visible from discovery through decision.

What it does

Intelligence with an inspectable foundation.

01 / Find context

Semantic AI search

Qdrant vector retrieval surfaces relevant passages from résumé and opportunity documents, even when the wording differs. Those passages provide source context for RAG analysis and tailored materials.

02 / Apply criteria

Deterministic logic

Explicit role criteria, résumé evidence, and saved career goals drive repeatable fit scoring. Optional AI analysis structures requirements and adds readable notes; the underlying criteria remain reviewable.

03 / Discover

Opportunity pipeline

Bring roles together from configured employer boards, job APIs, and selected search sources, then move them through review and workflow stages.

04 / Connect

CRM and planning

Link companies, people, networking history, interviews, meetings, and dated follow-ups to the opportunities that matter.

05 / Assist

MCP integrations

Compatible AI clients can search and review opportunities, read CRM context, run scoring workflows, and prepare scheduling or application tasks.

06 / Review

Human-led action

Recommendations can be inspected and corrected before they shape the workflow. Messaging tools prepare and track outreach; messages are sent personally.

How it works

From signal to informed action.

  1. CollectGather and organize opportunities from configured sources.
  2. RetrieveUse semantic search to find relevant résumé and role evidence.
  3. EvaluateApply deterministic criteria, with optional AI fit analysis.
  4. Follow throughReview the result, connect people and events, and plan the next action.
JobsAI calendar with fictional networking, interview, meeting, and follow-up events.
Calendar view with fictional sample events and contacts.

Behind the workspace

Built for a private, connected workflow.

Evidence and retrieval

Qdrant supports vector retrieval for résumé and opportunity passages. PostgreSQL stores opportunity, CRM, calendar, and workflow data.

Interfaces and processing

Containerized web and worker services support the dashboard and processing pipeline. An MCP server exposes selected tools to compatible assistants.

Hosting and recovery

A NAS-hosted production environment uses PostgreSQL as the primary database, with OCI and AWS standby environments. Database replication and retrieval data have separate recovery requirements.

This is a high-level view of a private implementation. Screenshots use fictional names, organizations, and activity.