Hephaestus
Ingestion & Profiling
1 Overview
D Discovery Review
2 Structure Map
P Ingestion Plan
3 Source Modules
C Coverage Census
4 Capability Profile
5 Qualitative Context
Analysis Setup
6 Audit Log
6b Source Graph
7 PM Profiles
8 Lab Run Setup
Hypothesis & Testing
9 Investment Thesis
10 Missing Data
11 Problem Solved
12 Hypothesis Board
Evaluation
13 Kill Mode
14 Evaluation Router
15 Remediation Tasks
16 Eval Monitor
Delivery
17 Pack Preview
18 Human Review
Governance
19 Audit Board

Overview

Dataset registration details and profiling status

Registration Details

Loading...

Profiling

No profiling runs yet.

Discovery Review

AI-assisted folder discovery: review, approve, reject, re-search

Keywords / Search Hints

Optional keywords, aliases, source names, regions, languages, tickers or concepts that help Hephaestus discover relevant folders.

Helps the AI focus LLM reranking on the most relevant candidates. Leave blank to use default (top 40).

Source Locations

Exact locations discovery will use instead of scanning the whole root. Edit and save before re-running discovery.

Structure Map

Discovered data locations and folder structures

Run profiling to discover data locations.

Ingestion Plan

Strategy-driven plan for how the dataset is profiled

Run profiling to generate an ingestion plan.

Source Modules

Individual data source profiles discovered from S3

Run profiling to discover source modules.

Coverage Census

Path-level entity and date coverage analysis from S3 key structure

Run coverage census to analyze entity/date coverage from S3 paths.

Capability Profile

Aggregate intelligence capabilities across all source modules

Run profiling to generate capability profile.

Qualitative Context

Evidence-grounded descriptions of dataset capabilities

Run profiling to generate qualitative context.

Audit Log

Immutable record of all actions taken on this dataset

No audit entries yet.

Implementation Audit Board

Track assumptions, tech debt, failures, and decisions

No audit items yet.

Source Graph & Capability Registry

What data Hermes already owns — LLM-classified assets, capabilities, and semantic columns. The source of truth for dependency resolution and new product ideas.

Data Estate Audit

Loading…

Run an audit to populate the Source Graph.

PM Profiles

Portfolio manager mandate profiles for thesis generation

Create PM Profile

Existing Profiles

Lab Run Setup

Configure and launch mandate-fit reasoning experiments

Launch Lab Run

Past Lab Runs

Investment Thesis

Thesis candidates generated from PM mandate + dataset capability

No lab run completed yet. Launch a lab run from the Lab Run Setup panel.

Missing Data

Data dependencies that are not yet available

No missing data analysis yet. Complete a lab run first.

Problem Solved

Submit data resolutions for missing data items

No missing data items to resolve. Complete a lab run first.

Hypothesis Board

Summary dashboard of thesis generation and lab run results

No lab runs completed yet.

Kill Mode

Systematic evaluation for dataset termination decisions. Coming in Build Block 4.

Evaluation Router

Route evaluation tasks to appropriate engines. Coming in Build Block 4.

Remediation Tasks

Automated remediation for data quality issues. Coming in Build Block 4.

Evaluation Monitor

Real-time monitoring of evaluation pipeline. Coming in Build Block 4.

Pack Preview

Preview deliverable intelligence packs. Coming in Build Block 5.

Human Review

Human-in-the-loop review before delivery. Coming in Build Block 5.