ACHIEVEMENT UNLOCKED
KONAMI INITIATE — you know the old ways.

PLAYER 1 — UNITED KINGDOM — UTC+0

RANNE GERODIAS

DEVELOPER // DESIGNER // CREATOR

“Build tools, not just features.”

Full-stack engineer with a DevOps and data background — cloud delivery at scale, enterprise IT, and analytics. Now building local-first, privacy-aware products: TypeScript/React and Tauri up front, Python + RAG/Ollama behind, MCP and agent tooling where it earns its place.

CAFFEINATED MOTIVATED CREATIVE ◉ AVAILABLE FOR HIRE
01 / CHARACTER SHEET

PLAYER STATS

player.dat
RANNE GERODIAS CLASS: ENGINEER · LV. MSc
STR
0
DEX
0
INT
0
CHA
0
  • Ships local-first surfaces where privacy and ownership are part of the product story.
  • Automates the boring path — infra, kits, pipelines — so teams stop firefighting the same week.
  • Grounds AI in tools you control: RAG on your files, MCP in the editor, agents on your own metal.
  • Keeps L2–L4 and endpoint reality in mind when something breaks in prod or on a user laptop.
02 / QUEST LOG

FEATURED QUESTS

QUEST 01 / 04
stitch.exe — running
LEGENDARY IN PROGRESS ST-001

STITCH

Own your subscription data — and search it locally.

Local-first subscription companion: Tauri + React desktop, Python Flask bridge for RAG over your PDFs, Google sign-in + Gmail discovery, optional face MFA. Not another siloed SaaS dashboard.

  • TAURI
  • REACT
  • PYTHON
  • FLASK
  • RAG
  • OAUTH
linkup_mcp — stdio
LEGENDARY IN PROGRESS LM-001

LINKUP MCP

AI assistance that stays on your machine, in your editor.

Custom MCP server for Cursor: Linkup web search plus local RAG over your documents (LlamaIndex + Ollama). Powers Stitch's document brain and editor-side tooling.

  • PYTHON
  • MCP
  • LLAMAINDEX
  • OLLAMA
  • FLASK
lucky-charm — winner
LEGENDARY COMPLETED LC-001

LUCKY CHARM

🏆 Encode Club hackathon winner.

Privacy-preserving standup tool: wallet-first auth, TEE-oriented processing, filtered dashboards — structured themes and actions without verbatim quotes leaving the trust boundary.

  • REACT
  • VITE
  • WALLET AUTH
  • TEE
  • FLASK
03 / INVENTORY

SKILL TREE

CLOUD & DEVOPS 0
AWS · MULTI-REGION
TERRAFORM / IaC
DOCKER / K8S / JENKINS
CI/CD · PYTHON / JAVA
AI & MCP 0
OLLAMA / LOCAL LLMS
LLAMAINDEX / RAG
MCP / CURSOR TOOLING
AGENT ORCHESTRATION
PRODUCT / STACK 0
TYPESCRIPT / REACT
TAURI / VITE
FLASK / PYTHON APIs
TAILWIND / LOCAL-FIRST UX
DATA & SCRIPTING 0
PYTHON
SQL
BASH
SAS
NETWORK & ENDPOINTS 0
WI-FI / LAN / DHCP / DNS
VoIP / CONFERENCING
WIN / MAC / LINUX
INTUNE / SNOW / AD / M365
PROCESS & CLIENT 0
AGILE / SCRUM / RUNBOOKS
STAKEHOLDER REPORTING
TEAM SCHEDULING & OPS
TECH WRITING / ITSM
04 / SAVE HISTORY

CAREER.LOG

C:\CAREER — VT323 terminal

C:\CAREER> cantium.exe --role "HELP DESK ENGINEER" [FEB 2026 — NOW]

First-line support across Windows/macOS/Linux — M365, VPN, security tooling. Hands-on repairs, network triage, VoIP and conference rooms. Logged in Intune, ServiceNow, AD.

C:\CAREER> dome.exe --role "MANAGER" [JUN 2025 — FEB 2026]

Recruited, trained and led a team; kept suppliers, stock and scheduling predictable through busy periods.

C:\CAREER> deloitte.exe --role "JUNIOR DEVOPS ENGINEER" [DEC 2021 — JUL 2023]

Automated multi-region AWS provisioning as IaC, wired into Jenkins with Python and Java — cut a long manual process to roughly a day. Agile delivery, docs and SOPs across AWS, Terraform, Docker, K8s.

C:\CAREER> greenstone.exe --role "JUNIOR PRODUCTION PROGRAMMER" [OCT 2019 — MAR 2020]

Automated database workflows with Bash, Python, SAS, SQL; helped migrate source control into GitHub.

C:\CAREER> katalyze.exe --role "DATA ANALYST" [JUL 2019 — OCT 2019]

Python for deployment automation and monitoring; designed and tuned SQL for faster retrieval.

C:\CAREER> dwp.exe --role "SUPPORT ANALYST" [DEC 2018 — MAY 2019]

Reports and visualisations so stakeholders could act on evidence; improved data collection practices.

C:\CAREER> type education.txt

UNIVERSITY OF KENT — COMPUTER SCIENCE MSc (2020–21) · MATHEMATICS BSc (2015–19)

C:\CAREER>

05 / SAVE POINT

SAVE YOUR PROGRESS?

Hiring for local-first product work, MCP/RAG tooling, or DevOps that actually removes toil? The save point is right here.