AI Systems Engineer · Agent Builder · Supply Chain AI

Borina Keo

Building intelligent agents across Supply Chain, Healthcare, and Retail — from ERP, WMS & TMS integration to autonomous AI systems that think, act, and remember.

15+ years bridging data science and enterprise operations. Researching LLM memory architectures and agent security while deploying AI across Kohl & Frisch and enterprise clients.

Supply Chain AIAgent EngineeringERP · WMS · TMSLLM Memory ResearchAgent SecurityHealthcare IntegrationRetail OperationsAI System DesignData SciencePython · MLAgentic WorkflowsClaude · GPT-4o · GeminiSupply Chain AIAgent EngineeringERP · WMS · TMSLLM Memory ResearchAgent SecurityHealthcare IntegrationRetail OperationsAI System DesignData SciencePython · MLAgentic WorkflowsClaude · GPT-4o · Gemini

Ask Illumi
(agent)

Type a question or pick a prompt — the AI answers using Borina's verified data.

Answer
Ask a question to see the response here.
Visual
Metrics
Chart

About

AI Systems
Engineer &
Agent Builder

I engineer AI systems that operate at enterprise scale — building autonomous agents with custom harnesses, skills, and workflows, grounded in 15+ years of supply chain, healthcare, and retail domain expertise.

My foundation is data science and business operations: I've led ERP, WMS, and TMS transformations across supply chain, healthcare distribution, and retail networks — understanding the messy, real-world data that AI models need to actually work. That operational depth is what separates systems that demo well from ones that survive production.

Now I focus on the frontier: designing agent architectures with harnesses, skills, and multi-step workflows; researching LLM memory systems (episodic, semantic, and working memory); and investigating agent security — prompt injection, trust boundaries, and authorization patterns for autonomous systems. The goal is AI that is not just capable, but reliable and safe.

Currently: AI & Agent Systems Lead · Kohl and Frisch
0+ Years in Enterprise AI & Ops
0+ Industries: Supply Chain, Healthcare, Retail
0+ Autonomous Agent Systems Deployed
Expertise

Skills &
Expertise

Specialized across AI Agents · Supply Chain · Data Science · Security

Career

Experience

Quantified impact across AI engineering, supply chain, and enterprise operations

Kohl and Frisch 2023 – Present

AI & Agent Systems Lead

  • Designed and deployed multi-step agentic workflows with custom skills and harness architecture, automating 16 operational processes across finance and supply chain
  • Built LLM-powered demand forecasting agents integrating ERP supply chain data, achieving 13% forecast accuracy improvement
  • Researching LLM memory patterns (episodic, semantic, working) and agent security threat models including prompt injection and privilege escalation
Kohl and Frisch 2019 – 2023

AI & ERP Integration Lead — Supply Chain Transformation

  • Directed end-to-end ERP modernization for Kohl & Frisch and enterprise supply chain clients, reducing operational costs by 30%
  • Integrated AI procurement agents into ERP workflows, reducing purchase cycle time by 33%
  • Led 45-person cross-functional team through 18-month platform migration and AI augmentation program
Kohl and Frisch 2016 – 2019

WMS & Operations AI Lead — Warehouse Intelligence

  • Deployed intelligent WMS across 9 distribution centers, integrating AI pick optimization and increasing throughput by 15%
  • Applied ML-driven operations models to reduce labor costs by 5% while improving pick accuracy to 99.7%
  • Partnered with C-suite to align warehouse AI roadmap with 5-year supply chain and retail fulfillment strategy
Work

AI Agent
Systems

16 Processes automated end-to-end

Autonomous Supply Chain Agent

Agent Harness · Skills · Workflows Claude / GPT-4o

Designed a multi-agent system with custom harness, domain skills, and multi-step planning workflows — autonomously managing demand signals, procurement triggers, and ERP updates across supply chain operations.

View on GitHub
13% Forecast accuracy improvement

LLM Memory Research — Supply Chain Context

Episodic · Semantic · Working Memory Python · LLM APIs

Researching LLM memory architectures (episodic recall, semantic knowledge graphs, working context windows) to ground supply chain AI agents in long-horizon operational history rather than single-session context.

100% Privilege escalation attempts blocked

Agent Security Research & Implementation

Prompt Injection · Trust Boundaries Red Teaming · Defense Patterns

Investigating threat models unique to autonomous agents: prompt injection via tool outputs, cross-agent trust escalation, and authorization boundaries. Developing defensive harness patterns that constrain agent actions to least-privilege scopes.

View on GitHub
Contact

Let's
build
something.

Open to consulting on AI agent systems, supply chain and healthcare integrations, LLM memory and security research, and full-time AI engineering roles.