From the listing
About the role
Invisible Technologies makes AI work. Our end-to-end AI platform structures messy data, automates digital workflows, deploys agentic solutions, measures outcomes, and integrates human expertise where it matters most. We have improved models for more than 80% of the world’s top AI companies, including Microsoft, AWS, and Cohere.
We're looking for a Director of Engineering to set the technical strategy for Data Labs, the team that builds the benchmarking systems, data pipelines, and RL environments behind our AI training engagements. You'll decide which capabilities we invest in, which we standardize into platform, and which we leave to individual engagements.
You'll lead a group of engineering managers, forward deployed engineers, and senior ICs, and partner closely with applied research, product, and go-to-market leadership. This is a hands-on strategic role where you'll make and defend hard architectural calls and stay close to clients to know which calls matter.
What you'll do
- Own the multi-year technical strategy for Data Labs and translate it into a roadmap tied to the AI training and evaluation market.
- Decide where to build shared platform versus engagement-specific solutions.
- Set architectural direction for benchmarking infrastructure, data pipelines, and RL environments.
- Anticipate technical shifts 6–18 months out, such as new post-training techniques, agentic evaluation, and synthetic data.
- Track the competitive landscape across AI data, evaluation, and RL environment providers.
- Define Invisible's point of view on benchmarking and evaluation, including capability gap analyses and regression tracking.
- Represent engineering in executive conversations, strategic client relationships, and pre-sales.
- Lead and grow an engineering org of multiple teams, including engineering managers and Staff+ ICs.
- Coach managers to build sound technical judgment and strategic awareness in their teams.
What they're looking for
- 10+ years in software engineering, with significant experience building ML or AI systems.
- 5+ years in engineering leadership, including experience managing managers across multiple teams.
- Track record of setting multi-year technical strategy and making platform investment decisions.
- Deep, current understanding of the AI training and evaluation landscape (fine-tuning, RLHF/RLAIF).
- Hands-on familiarity with LLM training pipelines.
- Experience building benchmarking, data curation, or RL infrastructure.
- Strong cross-functional partnership skills with research, product, and commercial teams.
- Excellent executive communication and confidence acting as a technical advisor.
Nice to have
- Prior experience at an AI services firm, applied research consultancy, or frontier lab.
- A visible point of view on the AI data and evaluation market (talks, writing, or open-source work).
- Familiarity with data licensing, synthetic data generation, or contamination detection.
- Experience shaping new product lines or service offerings from a technical foundation.
- Experience leading an org through rapid growth or significant change.
From the company page
Invisible benefits
◉ Remote policy · in their words
“We believe in recognizing exceptional work with exceptional benefits. That’s why we empower our Partners to work remotely around the world on a schedule that suits their lifestyle.”
Pulled from their careers page · verified by Atlas
The hot takeJeff's read on Invisible
Invisible's Glassdoor reviews are hovering around 3.5, which raises a bit of a yellow flag for me. Again, take GD reviews with a grain of salt, and simply an indicator. Here's a few things I DO like quite a bit. Check out their statement about autonomy and productivity on this page. Here's a few of their beliefs I align with:
👩💻 On productivity: "We believe work isn't about where you are or how many hours you log. It's about what you accomplish."
🏖️ On PTO: "Time off when you need it. No red tape. We care about outcomes, not optics."
🤓 On flexible hours: "Flexible work hours. Your hours are yours. What matters is the result, not when you open your laptop."