ANAKATA

Strategic Consulting

Bangkok, Thailand

ANAKATA
DriLAb

Our Platform

AI-Powered Competency Development

Our platform automates the entire journey from competency assessment to personalized learning. A multi-agent AI pipeline delivers results in minutes, not weeks.

How It Works

Three Integrated Systems

Our platform consists of three interconnected systems that work together to create a complete competency development lifecycle — from assessment through AI analysis to personalized learning.

01

Assess

System A — Competency Evaluation

Employees complete a structured competency assessment aligned to their position requirements. Scores are mapped against a Position-Competency Matrix to quantify skill levels across all relevant dimensions.

Position-Competency Matrix

Define required competency levels for every position in your organization. The matrix serves as the benchmark against which all assessments are measured.

Self & Admin Assessment

Employees can self-assess, or administrators can conduct assessments directly. Both modes produce quantitative scores on a 1-5 scale across all relevant competencies.

Configurable Frameworks

Create and customize competency frameworks that match your organization's specific needs — from AI foundations to strategic leadership capabilities.

Gap Identification

Automatically calculate the difference between required and actual competency levels. Visual gap analysis highlights priority areas for development.

02

Analyze & Design

System C — Multi-Agent AI Pipeline

Four specialized AI agents work in sequence: the System Analyst identifies gaps, the Learning Designer creates modules, the Resource Researcher matches courses via RAG search, and the Project Manager validates the final plan.

System Analyst Agent

Analyzes assessment scores against position requirements, identifies and prioritizes skill gaps, and generates a structured gap report for the downstream agents.

Learning Designer Agent

Takes the gap analysis and designs tailored learning modules — each with specific objectives, recommended duration, difficulty level, and learning outcomes.

Resource Researcher Agent

Uses semantic search (RAG) against a ChromaDB vector database of courses to find the most relevant learning resources for each module.

Project Manager Agent

Orchestrates the entire pipeline, validates outputs at each stage, assembles the final learning path, and saves results to the database.

03

Learn & Grow

System B — Learning Platform

Each employee receives a personalized learning path with curated courses, progress tracking, and clear milestones. The platform closes the loop between assessment, AI analysis, and actual skill development.

Personalized Learning Paths

Every learning path is unique — generated specifically for the individual based on their assessment results, position requirements, and skill gaps.

Curated Course Matching

Courses are matched using semantic similarity via vector embeddings, ensuring relevance goes beyond keyword matching to true content alignment.

Progress Dashboard

Track learning progress across all assigned modules and courses. Visual indicators show completion status, time invested, and competency improvement.

Continuous Improvement

Reassess periodically to measure competency growth. The platform tracks improvement over time and adjusts learning paths as skills develop.

Under the Hood

Multi-Agent AI Pipeline

Four specialized AI agents collaborate in a linear relay workflow, each contributing their expertise to produce a comprehensive, personalized learning plan.

1

Project Manager

Orchestrate & Validate

Initiates the pipeline, passes context to each agent, validates intermediate outputs, and assembles the final learning path. Acts as quality gate.

2

System Analyst

Identify Skill Gaps

Analyzes assessment scores against position requirements, calculates gap severity, and produces a prioritized list of development areas.

3

Learning Designer

Design Modules

Creates structured learning modules with objectives, difficulty levels, estimated duration, and clear outcomes for each identified gap.

4

Resource Researcher

Match Courses (RAG)

Searches the vector database using semantic similarity to find the best courses for each module. Returns ranked recommendations with relevance scores.

Technology Stack

LangChain & LangGraph

Agent Framework

Orchestrates multi-agent workflows with state management, conditional routing, and tool integration.

ChromaDB

Vector Database

Stores course embeddings for semantic similarity search. BGE-M3 embeddings enable multilingual course matching.

Ollama + Cloud APIs

LLM Providers

Local models for fast tasks, cloud APIs for complex reasoning. Automatic fallback ensures reliability.

PostgreSQL

Relational Database

Stores all structured data — users, assessments, competencies, learning paths, and progress tracking.

Experience It Yourself

Start your AI competency assessment and see how our multi-agent pipeline creates a personalized learning path in minutes.