# How to Start Freelancing as an Engineer (Contracts vs Projects)

## Executive summary

The video outlines the optimal freelance model for engineers, arguing that success requires combining long-term contracts (for stability and high revenue) with smaller, project-based work, and establishing recurring maintenance retainers. For build engineers, the focus should be on leveraging AI for back-office automation in small to medium-sized enterprises (SMBs) and developing proprietary software to build intellectual property (IP).

## Key takeaways

- Two Freelance Models: Long-term contracts are similar to full-time roles (AI/Software/Data Engineering), paid hourly/day rate, and typically secured via recruiters. Project-based work involves fixed-price, smaller scopes (e.g., building an AI assistant for a fixed price over two weeks).
- Revenue Source Focus: The speaker claims that 80% of revenue for full-time freelance tech professionals comes from long-term contracts and establishing relationships with enterprises, not from small, single-project gigs.
- Targeting SMBs for Projects: For solo project-based work, focus on small to medium-sized enterprises (SMBs) (up to ~200 employees). The opportunity lies in automating 'boring stuff' that every company needs, such as invoices, emails, data entry, and reporting.
- The Optimal Freelance Strategy: The most successful freelancers combine long-term contracts, stack smaller projects on top, and implement maintenance retainers. This combination allows for continuous revenue and the development of proprietary software (IP).

## Technical details

- AI/ML Engineering Roles: In 2026, high-demand roles include AI Engineering (integrating LLM features into existing products), building pipelines that feed AI, and advanced concepts like Retrieval-Augmented Generation (RAG), multi-agent systems, and evaluation.
- Back-Office Automation: This involves automating digital artifacts (invoices, emails, documents, phone calls) that flow through a company. Solutions can be trigger-based or webhook-based, utilizing assistants (e.g., Claude, Codex) connected to CRMs, ERPs, or ticketing systems.
- Implementation Architecture: Automation often requires a custom backend (e.g., Python) deployed on a Virtual Private Server (VPS) for background jobs. On-demand work can be solved by custom Multi-Channel Platform (MCP) servers and specialized skills.

## Practical implications

- To maximize freelance revenue, prioritize securing long-term contracts with enterprises while simultaneously building a portfolio of smaller, high-impact automation projects for SMBs.
- Focus technical efforts on back-office automation (e.g., connecting AI to existing CRMs/ERPs) rather than attempting to build entire, end-to-end enterprise systems solo.
- Establish maintenance retainers early in the client relationship. This provides predictable, recurring revenue that can fund the development of proprietary software (IP).

## Topics

Freelancing, AI Engineering, Software Development, Automation, Business Strategy, Build Engineering, Glaido Pro, Datalumina, Free Masterclass: Freelancing in the Age of Coding Agents

Source: https://www.youtube.com/watch?v=1HtydHmdhrQ