Founder · Repath ‘Ray’ Khan

One founder.
Many models.
One mission.

Repath Khan — publicly credited as “Ray” on most posts — is the Waterford-based founder of the Leemer Group. He started LeemerChat in 2023 as a resilience layer against model outages. Three years later it is a full workspace, an infrastructure company, a custom-model foundry, an agentic platform, and a code review product.

1B+

Tokens shipped

2023

First shipped

7+

Products live

Waterford

HQ

Short bio

Repath studied at the University of Limerick and founded Repath Enterprises Ltd before incubating LeemerChat. He is deliberately an Ireland-based builder: LeemerLabs operates out of Waterford with failover in Dublin, and almost all engineering decisions are made with European compliance posture as a non-negotiable input.

He is not a “prompt engineer” founder. Since LLaMA-1 he has fine-tuned Qwen, LLaMA, Gemma, Mistral, Mixtral, and a list of smaller models; built bilingual Bengali/English distillations; and designed orchestrator → worker chains inside LeemerChat before agents were fashionable. He tends to talk about models from the inside.

The through-line across everything he ships is the same argument: a single opaque API is the wrong abstraction for serious work, coordination beats raw scale, and open weights inside a European stack are how real organisations should operate. LeemerLabs is that argument, made as infrastructure.

In his own words

“The future of AI is not a single super-intelligence. It is systems that coordinate intelligence well.”
— KingLeemer launch, Feb 2026
“If an AI system cannot reason reliably without tools, giving it tools will not make it safer. It will just make it louder.”
— KingLeemer V1 rationale
“Your data. Your model. Our GPUs. Built in Waterford, scaling globally.”
— Foundry launch, Nov 2025
“We eat our own cooking. Every system we offer a client, we've already built for ourselves.”
— Foundry Report 2026

What he's built

Seven shipped products.

Not a roadmap. Not a pitch deck. A list of products that have processed real user traffic — typically before the rest of the market knew they were supposed to exist.

01

LeemerChat

2023 — present

The flagship multi-model workspace. Over 1B tokens processed through real user traffic. Codebase chat, background agents, PDF/research flows, 50+ models.

02

LeemerLabs

2025 — present

The research and infrastructure arm. Hosts the public inference gateway, Model Foundry, and the engineering behind everything the Leemer Group ships.

03

KingLeemer

2026

Multi-model consensus protocol. V1 reasoning council, V2 autonomous tool loops. The argument that coordination beats a bigger model.

04

Mission Control

2026

Agentic execution platform. "Drop a mission. Get intelligence." Cortex orchestrator, sandboxed workers, artifact-first output.

05

Critique.sh

2026

GitHub-native AI code review built on a sandbox-native architecture. Newest product in the Group.

06

Foundry

2025 — present

Custom LLM studio. Ireland's first Tinker partner. Fine-tunes 1B → 235B MoE models with exportable weights.

07

Warren.wiki · ExamMate · HeyCouncil · DeepThis

2024 — present

Domain-specific deployments in finance, education, civic, and research — built on the same infrastructure sold to clients.

Timeline

Shipping since 2023.

2023

LeemerChat starts as a resilience layer against model outages — a backup for when GPT-4 went down. Turns into something bigger.

2024

Workspace era. Research mode, writer, podcast flows, sharing. The product becomes a daily driver.

2025

Sprint year. v4.5 research and automation, v4.8 durable generation, LeemerLite on Groq, v4.9 PowerCode + LeemerGLM. November: LeemerLabs Model Foundry launches. Ireland's first Tinker beta partner.

2026

KingLeemer (Feb). Blackbox + Cursor Cloud integrations (Feb). Tinker goes GA, Foundry Report publishes (Feb). Mission Control (Mar). Critique.sh releases v3.1 on a 48-hour cadence. LeemerLabs opens its public inference gateway.

A Day With Ray

Direct access, no agency layer.

A focused session for SMEs and solo founders who need a senior-level AI strategy conversation without enterprise pricing. Same expertise that shapes LeemerLabs; tailored to one problem at a time.

90 MIN

Focused on one real problem

FOUNDER

No account managers, no handoffs

FOLLOW-UP

Written notes + concrete next steps