G2

Bilal M.

Research and Development Engineer

Medical Devices

Small-Business (50 or fewer emp.)

9/2/2026

"Huge Time-Saver for PR Reviews, but Droids Struggle with Complex Tasks"

4/5

Describe the project or task Factory AI helped with:

Factory AI offers a suite of tools designed to enhance software development workflows. Its flagship feature, the autonomous Droids, are AI-driven agents capable of handling various engineering tasks. These tasks range from code refactoring and maintenance to complex problem-solving and code review. By integrating seamlessly with platforms like GitHub, Factory AI aims to reduce developer fatigue and increase productivity by automating repetitive and time-consuming tasks. While it excels in certain areas, users should be aware of its limitations, particularly in handling tasks that require deep domain knowledge. Review collected by and hosted on G2.com.

What do you like best about Factory AI?

I’ve been using Factory.ai to streamline my development workflow, and their automated Code Review Droid has genuinely been a huge time-saver. Rather than giving vague or generic AI suggestions, it plugs directly into GitHub and reviews pull requests with real context. When I push a branch, the Droid immediately scans the changes, flags edge cases, catches potential bugs, and leaves clear inline comments exactly where they matter. What I especially appreciate is that it doesn’t create pointless noise; it stays focused on correctness and clean code, almost like having a dependable pair-programmer quietly checking my work in the background before anything merges. Overall, it’s simple, runs smoothly, and frees up a lot of mental energy so I can focus on building instead of second-guessing routine checks. When it is worth every penny: If you treat it like an automated junior engineer. If a single Droid refactors a 15-file migration, writes unit tests, or fixes a broken pull request autonomously, it easily saves 3–5 hours of senior engineering time in one go. Saving just two hours of developer time a month makes a $20 plan pay for itself many times over.

When it is NOT worth it: If you only need quick inline code autocompletions or basic chat help. Standard Copilot-style tools or free extensions are more than enough for simple code suggestions Review collected by and hosted on G2.com.

What do you dislike about Factory AI?

I tried Factory AI and its autonomous Droids for software development. The concept sounds great on paper, but the execution has some serious flaws. The biggest issue, by far, is how it handles complex tasks and overall code reliability. When I assign a Droid a multi-step task, it often gets stuck in loops, burns through a massive number of tokens, and produces half-baked or messy code that doesn’t follow standard architecture. Instead of saving me time, it turns into an “endless babysitting” situation: I end up spending more effort refactoring its files, fixing syntax bugs, and guiding it line by line than I would if I just wrote the script myself. For simple boilerplate, it’s fine, but the lack of domain context combined with unpredictable token drain makes it hard to trust for real engineering workflows. Review collected by and hosted on G2.com.

Recommendations to others considering Factory AI:

To maximize the benefits of Factory AI, it's crucial to understand its strengths and limitations. For tasks that require deep domain knowledge or complex problem-solving, it's best to use it as a supplementary tool rather than a primary solution. Pairing it with human oversight can ensure that the output meets the necessary quality standards. Additionally, regularly reviewing and updating the AI's training data can help improve its performance over time. Review collected by and hosted on G2.com.

What problems is Factory AI solving and how is that benefiting you?

Having used Factory AI in our engineering workflows, the standout feature for me is its autonomous agents, which they call Droids.

The biggest problem it solves for us is developer fatigue from multi-file refactoring, ongoing maintenance, and pull requests. Most coding AI tools just sit inside your code editor and offer line-by-line autocomplete, which still leaves the manual heavy lifting on you. With Factory’s Droids, I can hand over a higher-level engineering task, like upgrading deprecated libraries, writing test suites across modules, or fixing broken CI checks and let the agent inspect the whole codebase, propose a plan, and carry out the edits autonomously. It cuts down hours of tedious, repetitive boilerplate work and reduces the drag of code reviews, giving us more time to focus on core system design and critical data logic. You still need to review its diffs before merging (as you would with any good peer review), but having an automated “extra engineer” handling repo-wide chores has noticeably boosted our delivery speed. Review collected by and hosted on G2.com.