Aidena

— AI STACK RECOMMENDATION

AI Performance Analytics & Coaching Platform

End-to-end platform for ingesting employee data, analyzing performance patterns with LLMs, and generating personalized coaching recommendations at scale.

Stays alive for 365 days after the last visit.

Data & Analytics

AI Performance Analytics & Coaching Platform

End-to-end platform for ingesting employee data, analyzing performance patterns with LLMs, and generating personalized coaching recommendations at scale.

high confidence

Core Stack ℹ︎

Airbyte

Primary

Ingest employee data from HR systems, surveys, and performance tools into a centralized warehouse. 300+ connectors enable connecting to ATS, HRIS, and productivity platforms without custom ETL code.

$0-500/month

dbt

Primary

Transform raw employee data into clean, documented feature tables for analysis. SQL-first approach enables building performance metrics, tenure calculations, and coaching-relevant aggregations reproducibly.

$0-100/month

Claude Opus 4

Primary

Generate nuanced, contextual coaching recommendations from performance data. Opus handles complex reasoning about employee strengths, development areas, and personalized growth paths with low hallucination.

$50-300/month

Complete the Stack ℹ︎

CrewAI

Alternative

Orchestrate multi-agent workflows for performance analysis: one agent analyzes metrics, another generates coaching insights, another personalizes recommendations. Scales to batch-process hundreds of employees.

$0/month

Braintrust

Alternative

Evaluate quality of coaching recommendations against real employee outcomes. Build datasets of good/bad recommendations, run automated evals, and iterate on prompt quality continuously.

$0-200/month

Arize Phoenix

Alternative

Monitor LLM outputs in production. Track recommendation quality, latency, and cost. Debug failures when coaching suggestions miss the mark or recommendations are inconsistent.

$0/month

Getting started

  1. 1Set up Airbyte to sync employee data from your HRIS (Workday, BambooHR, etc.) and performance tools (Lattice, 15Five) into a data warehouse (Snowflake, BigQuery, Postgres).
  2. 2Use dbt to model performance metrics: engagement scores, goal completion rates, peer feedback aggregations, tenure, promotion readiness.
  3. 3Build a CrewAI workflow with agents: (a) Performance Analyzer agent queries dbt models and summarizes key metrics, (b) Coaching Agent uses Claude Opus to generate personalized recommendations, (c) Personalization Agent tailors advice to role/level.
  4. 4Integrate Braintrust to evaluate recommendation quality—collect feedback from managers on usefulness and iterate prompts.
  5. 5Deploy via Airflow or Dagster to run nightly analysis on all employees.
  6. 6Use Arize Phoenix to monitor LLM call quality, token usage, and latency in production.
  7. 7Expose results via API or dashboard for HR teams to review and act on recommendations.
Copy link to clipboard

What are you building?

Build your own AI stack →