Gorkem Ercan
@GorkemErcan
Followers
788
Following
300
Media
77
Statuses
3K
Doer of Deeds. CTO of Jozu, Distinguished Engineer, ex-Red Hat
Toronto, Ontario
Joined April 2009
Monorepos for AI Projects: The Good, the Bad, and the Ugly
gorkem-ercan.com
Recently, I met a team that used an AI‑focused monorepo containing everything: notebooks, training pipelines, microservices, and infrastructure code.
0
0
0
Between them, Red Hat and IBM lost more IDE talent than the entire AI IDE sector. Today there is a small core left to upkeep a few IDE plugins. What do you think the management lesson here?.
1
0
1
Generate videos in just a few seconds. Try Grok Imagine, free for a limited time.
766
3K
10K
Model Lifecycle ≠ Container Lifecycle — Why Jozu Splits Them
gorkem-ercan.com
Why prebuilt inference containers like NVIDIA NIMs miss the mark on lifecycle, control, and separation of concerns.
0
0
0
Why Model Packaging Isn’t Enough
gorkem-ercan.com
Docker’s recent release of Docker Model Runner sends a clear signal: AI developers want to package models like they package code.
0
0
0
From Dev to Deploy: Why We Package AI/ML Artifacts with OCI, Not Git
0
2
2
👉 Join my workshop at PlatformCon on June 26 at 11 AM EDT: "Trusting Your AI Models: Building a Secure Cloud-Native Supply Chain".🔗
0
0
0
3️⃣ Reproducible.Each ModelKit captures dependencies and configuration, ensuring consistent reproducability of AI/ML projects.
1
0
0
2️⃣ Built-in provenance.ModelKits embeds metadata right in the manifest—no external DB needed.
1
0
0
1️⃣ Immutable artifacts.Every ModelKit is content-addressed (digest-pinned), so you always know exactly what code, data & weights are used.
1
0
0
🔍 How critical is end-to-end provenance for your MCP servers?.📊 Vote below & share why in replies!.
0
0
0
Old but gold—packing LoRA adapters into ModelKits made easy. 🔧📦 Dive in: #MLOps.
jozu.com
Dive into the world of large language models with our step-by-step tutorial on fine-tuning using LoRA, powered by tools like llama.cpp and KitOps. LoRA (Low-Rank Adaptation) is an efficient technique...
0
0
1
RT @ramiyengar: I 💛 @Kit_Ops .I've been wanting to talk about the tool ever since I got their keycaps. My first chance was @KCDBengaluru. I….
0
4
0
ModelKits aren’t tied to Kit CLI. Just hit “Pull Tag” again and paste it to wherever you need it. It’s that simple. Your stack, your rules. We just make it seamless. 🔗 #MLOps #AIInfrastructure
0
2
3
Forget Notion. At Jozu, every doc starts as Markdown—and stays that way. We power reviews and comments in keeping our workflow lean, mean, and hackable. 🦾.
0
0
0
New KitOps release is out. KitOps 1.6.0 comes with bug fixes and a nifty new flag for optimizing unpacking models especially for users of initContainers.
github.com
Welcome to the v1.6.0 release of Kit! We've added some exciting new features and improvements. New Features Ignore existing files when using kit unpack Previously, kit unpack worked in one of t...
0
0
1
Jozu Raises $4 Million Seed Round to Scale Enterprise AI Platform #kitops.
jozu.com
Funding Led by HalfCourt Capital with Participation from Mozilla Ventures, BrightSpark, AlleyCorp, Sentiero, and Union Bay Partners New York - May 8, 2025 - Jozu, the company building enterprise-gr...
0
1
7
RT @Kit_Ops: Did you know?. KitOps leverages existing OCI registries already used in your org. You get enterprise features like RBAC, audi….
0
2
0
RT @cloudnativeboy: @Kit_Ops CNCF Sandbox is an open-source packaging & versioning solution designed for speed, security, & consistency for….
0
2
0