---
title: "Steriani Karamanlis (@steriani) | Hashnode"
description: "Co-founder & CMO at ATOM. Building the global price benchmark for AI inference. Read the latest articles by Steriani Karamanlis on Hashnode."
url: "https://hashnode.com/@steriani"
---

@steriani

# Steriani Karamanlis

@steriani · Athens, Greece · Joined March 2026

Co-founder & CMO at ATOM. Building the global price benchmark for AI inference.

## Comments

[One thing that does not get enough attention in LLM backend security discussions is how vendor diversity creates new attack surfaces. Most production systems now route across multiple inference providers depending on cost, latency and availability. Each of those providers has different authentication patterns, rate limiting behaviors and response formats. A secure by design approach has to account for the fact that the backend is not a single endpoint anymore but a dynamic mix of 50+ potential vendors depending on what is cheapest and fastest at any given moment. We track that vendor landscape weekly at a7om.com and the fragmentation is real.](https://hashnode.com/posts/secure-by-design-patterns-for-llm-backend-apis/69d76c88fa7251682e73447c/comment/69d92078c8e5007ddbc374ae)

Comment · Article · Apr 10 · 1 · [Secure-by-Design Patterns for LLM-Backend APIs](https://hashnode.com/posts/secure-by-design-patterns-for-llm-backend-apis/69d76c88fa7251682e73447c)

[The cost dimension of RAG failures in production is something we see reflected in the pricing data and it is underappreciated. Most teams prototype RAG with a retrieval step that pulls generous context windows to make sure nothing gets missed. Then they hit production and realize that output tokens cost roughly 4x input tokens on average across the market right now. The "retrieve more to be safe" instinct that worked in testing becomes an expensive habit at scale. The systems that survive production are usually the ones that got ruthless about retrieval precision early, not because of latency but because of the inference bill. We track inference costs across 50+ vendors weekly at a7om.com if the numbers are useful.](https://hashnode.com/posts/why-most-rag-systems-fail-in-production-and-how-to-fix-them/69d76192fa7251682e67c431/comment/69d92060c8e5007ddbc35f3d)

Comment · Article · Apr 10 · [Why Most RAG Systems Fail in Production (And How to Fix Them)](https://hashnode.com/posts/why-most-rag-systems-fail-in-production-and-how-to-fix-them/69d76192fa7251682e67c431)