Marathon Venture Capital

Excited to share our investment in @NOFireAI ⛔️🔥 Code generation volume has gone through the roof with AI - but all hell breaks loose when you deploy such code into production... NoFire AI builds the Production Context Graph – a live, time-versioned map of every service, dependency, deployment, configuration change and failure pattern with an organization's code base. It then leverages such to deliver its Context & Control Model, preventing and resolving incidents on the spot by tracing the actual causal chain from change to consequence, while operationalising all learnings along the way. The company brings together an observability all-star roster of peers that we are delighted to partner with: – Spiros Economakis, CEO, x-Head of Cloud at Mattermost & Lenses - Panagiotis Moustafellos, CTO, Distinguished Engineer & SRE Lead at Elastic - Anastassios Nanos, Chief Scientist, creator of urunc, the container runtime for unikernels - Antonis Chalkiopoulos, COO, founder of Lenses, acquired by Celonis At @marathon_vc we're excited to lead their $2.5m seed round. Here's a more personal note with further color on our investment by Panos Papadopoulos: marathon.vc/blog/no-err-twic… By mapping invisible system dependencies in real-time and red-flagging risky updates before deployment, NoFire AI pinpoints the exact root cause of complex failures and helps your engineering team to sleep well at night – make sure you give them a try!

Excited to share @marathon_vc investment in @straion_app – keeping your AI-generated code on track! We all love our coding agents, but their output is not aligned with your organisation’s rules, and the bottleneck has moved to code reviewing. The way the industry has been trying to solve this has turned to modern alchemy – sharing “hacks” and editing .md files here and there, with inconsistent results. Straion addresses the problem head on: a) You import your engineering / security / coding rules one-off, b) Straion dynamically updates the context window with only the relevant rules per task, c) Your agent’s output is validated against your rules and fixed on the spot, ready to ship without wasting time and tokens. Lukas Holzer (@luka5c0m), Katrin Freihofner and Fabian Friedl are Dynatrace & Elastic alumni and know a thing or two about the modern challenges of enterprise code compliance vs productivity. We are humbled to serve as their original investor, and look forward to Straion helping you make the best out of your AI agents starting today!

Excited to share @PallmaAI $1.7 million seed round led by @marathon_vc! Every major technology platform shift comes with brand new security challenges. The same applies to AI. Enterprises are rapidly adopting customer-facing agentic solutions. At the same time, they get exposed to unprecedented threats! Everything an AI agent has access to, might now also be accessible by malicious users. Prompt injections, data exfiltration, policy bypass attempts and unauthorised targets, to name a few, are being mentioned more and more frequently… @PavlosMitsoulis and Dionysis Varelas have built and deployed large-scale AI systems for a dozen years each at companies like Amazon, Expedia and King. Facing such challenges first-hand, paired with the lack of appropriate solutions by legacy security vendors, made them switch to founder mode and embark on a journey to build the native security platform for AI agents. Pallma AI offers enterprise-grade agent security, via proactive discovery and runtime enforcement. Their solution has been received with enthusiasm by household name customers, and they are just getting started! If you are excited about building state-of-the-art production systems at the intersection of AI and security, or want to protect your customer-facing AI infrastructure, reach out to the team! @alexalexakis #autonomous #agentic #AI #security