System Update // TM-01 Thinking Modes is now published. A practical framework for building intent-aware AI. View the Book →
Besh Identity
Get in Touch
AI Systems Architect & Engineer

Building AI that
acts with intent.

I design multi-agent systems and tools that execute with logic. Making AI reliable, traceable, and ready for real-world work.

> CORE_ENGINE: AGENTIC_V4
> STATUS: NOMINAL
> ROOTS: 25Y_FULL_STACK
> FOCUS: TRACEABILITY

Current Missions

Active Missions // Priority Queue ● Published

How to Think with AI: The Missing Skill. A book with a practical framework for mastering the 4 control knobs to get clear, useful answers every time.

Basa AI

An AI reading companion designed for low-resource settings to improve comprehension and critical thinking.

Tindlekit

An open platform for surfacing bold ideas, built using high-density AI collaboration.

Freelance Rate Calculator

Data-driven pricing engine for freelancers to bridge the global wage gap.

Terrakindle

NASA-powered planetary resilience engine and climate analysis.

Intelligence Reports

Signal Detection

View full archive →
Thinking Trace

Sketchnotes

View Archive →
Trace 71383175

Trace 71383175

Thinking Trace
Trace 88152058

Trace 88152058

Thinking Trace
Trace 71181320

Trace 71181320

Thinking Trace
Infrastructure

The Agentic Stack

Agent Workflows

Designing multi-agent orchestrations using Swarm and LangGraph for complex logic.

Prompt Architecture

Systematic engineering utilizing structured outputs and chain-of-thought verification.

Context Hygiene

Optimizing performance through rigorous context management and vector embedding.

Traceable Logic

Building AI systems that explain their reasoning, ensuring human-in-the-loop oversight.

Retrieval Strategy

Implementing vector search and semantic grounding to provide AI with reliable data foundations.

Human-in-the-loop

Designing seamless feedback interfaces where humans steer and validate autonomous agent actions.