ThalaAI
ThalaAI Applied AI Product Studio

Product / TalentEcho

TalentEcho

AI-assisted job intelligence across fragmented hiring systems.

TalentEcho is ThalaAI’s job intelligence product. It is built on a recall-first pipeline that aggregates roles from multiple ATS and company career systems, normalizes them, removes duplicates, and ranks them into a human-readable shortlist.

View Open Source Explore Architecture Contact
Early Recall-first engine ATS aggregation Ranked shortlist output

Architecture Overview

A multi-source pipeline built for high recall and practical review.

The current public engine ingests roles from a configurable YAML registry, runs source-specific adapters, normalizes and deduplicates the job stream, then ranks results into human-readable outputs with diagnostics for debugging coverage gaps.

Inputs Adapters Pipeline Outputs Diagnostics
Input + Interface

Candidate entry point

User / Candidate Search intent, company targets, role families, review workflow.
TalentEcho Web or CLI Experience Runs the open-source engine and packages a review-ready shortlist.
Source Registry YAML Companies + sources, metadata, enable / disable controls.
Adapter Layer

Multi-source adapters

Workday Greenhouse Lever Eightfold Amazon JSON Apple JSON-LD Microsoft Search Meta GraphQL
Core Pipeline

Recall-first processing flow

Raw Job Ingestion Collects structured roles broadly first.
Normalization Converts source payloads into a common job model.
Deduplication Removes overlaps across fragmented systems.
Ranking Engine Scores title, seniority, leadership, delivery, domain, and location fit.
Additional scoring input
Resume.txt Feeds future resume-aware scoring into ranking.
Outputs

Match outputs

  • job_matches.md
  • job_matches.csv
  • ranked_jobs.csv
  • normalized_jobs.csv
  • diagnostics.csv
Diagnostics Branch

Diagnostics / Source Health

Tracks raw, normalized, deduped, scored, and output counts plus source-level errors, making coverage gaps visible and debuggable.

How It Works

From fragmented ATS sources to a review-ready shortlist.

01

Multi-source ingestion

Collect jobs from ATS and company career systems through adapter-specific source families.

02

Normalize and dedupe

Convert source payloads into a common job format and remove overlapping postings.

03

Rank using candidate-fit signals

Score roles with title, seniority, leadership, domain, delivery, and location signals.

04

Review shortlist

Inspect human-readable markdown and CSV outputs for the strongest current matches.

05

Debug with diagnostics

Use diagnostics output to understand source health, stage counts, and ingestion gaps.

Product Highlights

Core capabilities in the current engine.

Recall-first ingestion

The engine is designed to collect broadly first, then narrow through normalization, deduplication, and ranking.

Configurable source registry

Companies and sources live in YAML, making coverage changes possible without editing Python just to add or disable a source.

Human-readable shortlist output

Results ship as markdown and CSV so a candidate can review matches quickly without needing a custom UI first.

Diagnostics and source health

Each run emits stage counts and error visibility so source failures are debuggable rather than hidden.

Built in the Open

The public engine already exists as open source.

The current public foundation for TalentEcho is the open-source `jobmatch-bot` / `job_intel` project. It collects, normalizes, deduplicates, diagnoses, and ranks jobs from structured ATS and company career sources.

Open GitHub Repository

Open Source Link

Community-facing foundation, product-facing direction.

TalentEcho uses the open-source engine as its factual base today while the broader product direction expands toward stronger ranking, wider coverage, and a future hosted experience.

Status / Roadmap

Early product with working ingestion and ranking foundations.

Early
  • Multi-source ingestion
  • Normalization
  • Deduplication
  • Ranking
  • Configurable source registry
  • Diagnostics

Next direction includes better ATS coverage, stronger normalization, resume-aware scoring, and a hosted version later.

Part of the ThalaAI Ecosystem

TalentEcho represents ThalaAI’s work in work intelligence.

Within the broader portfolio, TalentEcho is the work-intelligence and decision-support product line. It extends ThalaAI beyond wellness and creative systems into practical hiring and career signal workflows.

Back to Products Contact