Independent consultant · part of Bytecode

Rails rescue, DevOps, and practical AI

I help founders and product teams stabilize, modernize, and scale Ruby on Rails applications — then connect them to useful AI workflows.

I am a senior Ruby/Rails developer, DevOps engineer, and technical consultant from Zagreb, Croatia, with 15+ years of experience building production web applications, internal tools, data pipelines, and backend systems.

My best fit is remote contract work where an existing Rails product needs experienced hands: upgrades, refactoring, performance work, production debugging, infrastructure cleanup, payment/CRM integrations, or practical AI/LLM automation.

Available for: remote contract, freelance, part-time consulting, and CTO-for-hire engagements for founder-led SaaS products.

Start a conversation View selected work

15+

years of backend and Rails consulting

100k+

monthly users served by shipped products

111k

Rails training samples used for public Hugging Face models

250k+

records synchronized through production pipelines

How I Can Help

Senior help for Rails stability, speed, and AI leverage.

Ruby on Rails development, maintenance, upgrades, and refactoring
Rails rescue work: production debugging, performance fixes, test stabilization, and legacy cleanup
Hotwire, Turbo, Stimulus, PostgreSQL, Redis, Elastic, background jobs, and pragmatic full-stack Rails
DevOps, Docker, AWS, DigitalOcean, Heroku, CI/CD, monitoring, and server architecture
Stripe/payment integrations, CRM integrations, admin systems, and internal tools
AI/LLM integrations: internal copilots, data extraction, workflow automation, RAG-style search, and agent workflows
Technical partner / CTO-for-hire support for non-technical founders and small teams

If you have a Rails product that needs to be stabilized, modernized, scaled, or connected to AI/business automation workflows, email me at vmarcetic@bytecode.hr.

Selected Work

Production systems, rescue work, and practical AI automation.

Bytecode Rails LLMs

huggingface.co/bytecodehr

Public Hugging Face models fine-tuned for idiomatic Ruby on Rails code generation.

Problem: General coding models often miss team conventions and generate Rails code that looks plausible but does not match how mature Rails products are actually built and maintained.

What I did:

Built and published Bytecode’s Rails-focused Hugging Face models
Fine-tuned Qwen3-based models on 111,000 samples from internal Rails projects
Prepared GGUF/Ollama-friendly releases for local development workflows
Documented the dataset engineering, training, quantization, and deployment process

Result: Public Rails-specialized LLMs, including an 8B laptop-friendly model and a 31B MoE flagship model, that encode Bytecode’s Rails conventions directly into the weights.

Tech: Hugging Face, Qwen3, QLoRA, Unsloth, GGUF, Ollama, Ruby on Rails

8B Rails model 31B Rails model All Bytecode models

SayTrue

saytrue.ai

AI customer service platform that answers across chat, voice, and social channels only from approved knowledge, with source citations and human handoff when confidence is low.

Problem: Regulated teams need AI support agents that can explain where answers came from instead of hallucinating or improvising policy-sensitive responses.

What I did:

Built a source-grounded AI support platform for web chat, phone calls, Messenger, and Instagram DMs
Designed workflows around approved knowledge bases, citations, confidence thresholds, and honest fallback
Added voice/transcription flows and supervisor-friendly searchable transcripts
Shipped an EU-hosted, GDPR-aware product with an open changelog and demo/contact funnel

Tech: Ruby on Rails, AI/LLM workflows, RAG, voice transcription, omnichannel messaging, GDPR-aware infrastructure

Uskladi.eu

uskladi.eu

Compliance operating system for Croatian and EU organizations managing GDPR, NIS2, DORA, and EU AI Act obligations.

Problem: Compliance work usually lives across spreadsheets, documents, email threads, scans, and disconnected evidence folders, making audits and ongoing readiness painful.

What I did:

Built a platform for GAP assessments, readiness scoring, registers, findings, evidence, and audit exports
Modeled workflows for GDPR ROPA/DSR/DPIA, NIS2 incident timelines, DORA controls, and EU AI Act inventories
Connected technical scanning and compliance findings into one operational workspace
Localized the product for Croatian organizations and EU regulatory language

Tech: Ruby on Rails, PostgreSQL, compliance workflows, document exports, security scanning, EU regulatory frameworks

Napster

www.bytecode.hr/projects/napster

Web backend development and R&D support for one of the most recognizable brands in digital music.

Problem: A large, established streaming platform needed reliable backend contributions and fast technical exploration for a new product initiative.

What I did:

Contributed to web-facing backend systems inside an established product codebase
Supported R&D and prototyping for a new product direction
Worked within existing engineering standards and delivery workflows at scale

Tech: Ruby on Rails, PostgreSQL, Redis

Roomsie

www.bytecode.hr/projects/roomsie

Medium-term housing marketplace for interns, students, and recent graduates in Washington, D.C.

Problem: After multiple teams failed to deliver, the product needed a focused senior team to stabilize direction, rebuild, and launch.

What I did:

Took over an existing troubled build and helped move it toward a working product
Implemented payment flows, production features, and core marketplace functionality
Fixed and optimized Rails code and RSpec tests while keeping delivery moving

Result: Rebuilt and launched a revenue-generating product in 8 months.

Tech: Ruby on Rails, React, PostgreSQL, Redis, Heroku

RaceChip

www.bytecode.hr/projects/racechip

Back-end data system for a high-growth automotive ecommerce company selling performance tuning chips worldwide.

Problem: RaceChip needed accurate, scalable tooling for a vehicle/tuning data matrix covering thousands of models, with no room for incorrect tuning values.

What I did:

Worked on backend systems for a high-traffic commerce/product platform
Supported PostgreSQL-backed services and Java application development
Helped with AWS/Jenkins-based deployment and operational workflows

Tech: Java, Play, PostgreSQL, Redis, AWS, Jenkins

Movieo.me

movieo.me

Movie discovery platform aggregating ratings and metadata from IMDB, Rotten Tomatoes, Metacritic, TMDB, OMDB, and other sources. Featured on Hacker News, Product Hunt, and Fast Company; used by 100,000+ people per month at its peak.

Problem: People needed a fast, pleasant way to discover what to watch across a large movie catalog with fragmented data sources.

What I did:

Built automated ingestion/update pipelines for 250,000+ movies
Designed caching and data delivery with Elastic and Redis for a fast user experience
Implemented advanced filtering and recommendation flows for movie discovery
Kept external API data synchronized and production-ready through background jobs

Result: A high-traffic, widely shared product with strong community reception and reliable daily data refreshes.

Tech: Ruby on Rails, PostgreSQL, Elastic, Redis, Thumbor

Fast Company coverage

TwoMargins

www.bytecode.hr/projects/twomargins

Collaborative financial research platform for annotating SEC filings and other complex financial and political documents.

Problem: SEC filings are large, irregular HTML documents that are difficult to parse, annotate, and integrate into partner workflows.

What I did:

Built the Rails backend and deployed the production infrastructure from scratch
Created a Nokogiri-powered parser for deeply nested SEC HTML documents
Integrated the annotation interface into external partner platforms
Supported product owners during peak-traffic document releases and U.S. election cycles
Onboarded and trained developers as the project matured

Tech: Ruby on Rails, PostgreSQL, Nokogiri

Wall Street Journal coverage

Cora

www.bytecode.hr/projects/cora

Backend development support for an organic period and body care brand with ecommerce and subscription operations.

Problem: A growing D2C brand needed focused backend help inside an existing production product without disrupting existing workflows.

What I did:

Supported backend services behind product catalog, subscription management, and order processing
Stepped into an existing codebase for targeted delivery
Helped keep ecommerce operations moving for a mission-driven product team

Tech: Ruby on Rails, PostgreSQL