From Invisible Technologies's posting. “We” and “our” refer to the employer.
About Invisible
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.
Our platform cleans, labels, and structures company data so it is ready for AI. It adapts models to each business and adds human expertise when needed, the same approach we have used to improve models for more than 80% of the world’s top AI companies, including Microsoft, AWS, and Cohere.
Our successes span industries, from supply chain automation for Swiss Gear to AI-enabled naval simulations with SAIC, and validating NBA draft picks for the Charlotte Hornets.
Profitable for more than half a decade, Invisible reached $134M in revenue and ranked as the number two fastest growing AI company on the 2024 Inc. 5000. In September 2025, we raised $100M in growth capital to accelerate our mission of making AI actually work in the enterprise and to advance our platform technology.
About the Role
Invisible's Data Lab works with leading AI companies to design and deliver the human data that advances model capabilities and measures performance. Our work spans the full frontier of AI training and evaluation—from supervised fine-tuning and preference data to benchmarks, red teaming, multimodal data, coding tasks, and agent environments. Solutions architects are at the center of this work, working with clients to define the problem and translating open-ended model development problems into scalable programs with sound methodology.
We are looking for a Senior Solutions Architect to bring technical depth, sharp commercial judgment, and a strong command of the rapidly evolving AI training and evaluation landscape. Embedded within an account team working with a major frontier AI lab, you will lead solution design for new opportunities and pilots.
You will partner with Project Leads who own execution, client management, and overall project outcomes, while remaining accountable for the integrity, feasibility, and scalability of the solution through handoff and early delivery.
You will often enter when a client's ask is still underspecified. Your job is to understand what they are really trying to achieve, translate that into a collection we can execute, and help the client and our teams make the right technical, methodological, and commercial choices. You will connect the perspectives of client researchers, program managers, Project Leads, experts, and engineers—bringing context that may not otherwise exist in the room.
This is not a traditional infrastructure architecture role, nor is it primarily a project management role. It sits at the intersection of ML research, human data, product thinking, and client partnership. Success means teams make better decisions because you are involved: projects are framed correctly, solutions reflect the state of the art, and clients trust Invisible to understand where their work is going next.
You will work across a remarkable range of frontier AI problems, with direct exposure to the researchers and program leaders defining them. You will help determine how novel human-data programs should work before there is an established playbook, influence the products and systems used to deliver them, and build reusable ways of thinking that shape how Invisible approaches the field.
If you live and breathe LLM training and evaluation, communicate with unusual clarity, and want your judgment to matter in the room, this role offers a rare combination of technical range, client proximity, and real-world impact.
What You'll Do
Equip account teams with the technical and research context they need to fully understand client requests and make sound decisions about how to approach them.
Own the solution design of new human-data projects and pilots, working with Project Leads from early discovery and scoping through the transition into execution.
Work directly with client research and program management stakeholders to interpret ambiguous requests, uncover the underlying objective, and turn it into a clear and executable approach. For technical POCs in particular, you may take the lead in the relationship alongside the Project Lead.
Advise on methodology/best practices across supervised fine-tuning, preference data and RLHF, benchmarks and evaluations, red teaming, multimodal data, coding tasks, agent environments, and emerging approaches.
Shape task design, annotation and evaluation workflows, quality and validation strategies, data structures, and the trade-offs between research value, feasibility, speed, scale, and expert experience.
Define the expert profiles a project requires, including the knowledge, experience, and judgment needed to produce credible data.
Bring commercial judgment to early scoping, building rough-cut estimates of effort, expert supply, timeline, and cost so teams can test feasibility and shape viable proposals.
Partner with Engineering to define the expert-facing interfaces and technical architecture that collections require, including workflows, APIs, data exchange patterns, synthetic testing, and validation mechanisms.
Prototype or test critical elements of a solution during discovery, while working with Engineering as the team responsible for production implementation.
Build alignment across clients, account leadership, Project Leads, Engineering, expert operations, and other internal teams.
Stay deeply current on LLM training, post-training, human data, benchmarks, and evaluation research, and translate new developments into practical implications for active and future client work.
Develop and share reusable mental models, patterns, and emerging best practices that raise the technical fluency and solutioning quality of the wider organization.
What We Need
Required
Deep, demonstrable interest in how frontier AI models are trained, improved, and evaluated. You actively follow the space and can form your own views rather than simply repeat its vocabulary.
Strong working knowledge of LLM training and evaluation concepts, with the ability to reason across data modalities, collection methods, and research objectives.
Exceptional communication skills. You can listen closely, identify what is missing from an underspecified request, explain complex ideas clearly, and influence technical and non-technical stakeholders.
The judgment to challenge assumptions constructively while maintaining momentum and client trust. You know when to ask another question, when to recommend a different approach, and when to make a pragmatic call.
Enough technical fluency in ML, data systems, software architecture, APIs, and data flows to collaborate credibly with researchers and engineers and to shape implementable solutions.
Commercial judgment and comfort with rough-cut budgeting, particularly when weighing research ambition against expert availability, delivery effort, timeline, cost, and unit economics.
Ability to synthesize client context, research intent, expert capabilities, user experience, technical constraints, and operational realities into an executable approach.
Experience working across multiple stakeholders or teams, particularly in ambiguous environments where responsibilities and requirements continuously evolve.
High ownership, curiosity, adaptability, and low ego. You care more about reaching the right answer and improving the outcome than being the person who supplied it.
Nice to Have
A background in computer science, software engineering, machine learning, data science, AI research, technical product, solutions architecture, or a related field. Equivalent knowledge gained through independent work or other professional experience is equally welcome.
Direct experience with human-data programs, data annotation, model post-training, RLHF, RLVR, evaluations, benchmarks, red teaming, or research operations.
Experience partnering directly with AI researchers, research engineers, or technical program managers.
Hands-on ability with tools such as Python, SQL, APIs, notebooks, or lightweight prototyping. You will not be the primary builder, but being able to inspect, test, or demonstrate an idea can make you more effective.
Experience designing tools or workflows for specialized users, particularly where expert judgment and data quality are central to the outcome.
Evidence of serious engagement with the field through professional work, research, open-source contributions, technical writing, independent projects, benchmark participation, or other forms of self-directed learning.
You can find more information about our geographic pay tiers here.
During the interview process, your Invisible Talent Acquisition Partner will confirm which tier applies to your location. For candidates outside the U.S., compensation is adjusted to reflect local market conditions and cost of living.
*Bonuses and equity are included in all full-time offers. Final compensation is determined by a combination of factors, including location, job-related experience, skills, knowledge, internal pay equity, and overall market conditions. Because of this, every offer is unique. Additional details on total compensation and benefits will be discussed during the hiring process.
What It's Like to Work at Invisible:
At Invisible, we’re not just redefining work—we’re reinventing it. We operate at the intersection of advanced AI and human ingenuity, pushing the boundaries of what’s possible to unlock productivity and scale. Ownership is at the core of everything we do. Here, you won’t just execute tasks—you’ll build, innovate, and shape the future alongside world-class clients pushing the boundaries of AI.
We expect bold ideas, relentless drive, and the ability to turn ambiguity into opportunity. The pace is fast, the challenges are big, and the growth is unmatched. We’re not for everyone, and we’re okay with that. If you’re looking for predictable routines, this isn’t the place for you. But if you’re driven to create, thrive in dynamic environments, and want a front-row seat to the AI revolution, you’ll fit right in.
Country Hiring Guidelines:
Invisible is a hybrid organization with offices and team members located around the world. While some roles may offer remote flexibility, most positions involve in-office collaboration and are tied to specific locations. Any location-based requirements or hybrid expectations will be communicated by our Talent Acquisition team during the recruiting process.
AI Interviewing Guidelines:
Our hiring team thoughtfully uses AI to support an efficient, engaging, and inclusive interview process. Since AI can also be a helpful tool for candidates, we've outlined expectations for using it ethically throughout your interview journey.
Click here to learn more about how we use AI and our guidelines for candidates.
Accessibility Statement:
We are committed to providing reasonable accommodations for individuals with disabilities. If you require an accommodation to participate in the application or interview process, please submit your request using our accommodation request form. A member of our team will follow up to support your request.
Here is the link the google form: https://docs.google.com/forms/d/e/1FAIpQLScqsu0PCQ0m0lwt-VFeXovNbZxWk0jogv8QI_ciGqZsIbSBjQ/viewform?usp=sharing&ouid=102356854798492362557
Equal Opportunity Statement:
We’re an equal opportunity employer. All qualified applicants will receive consideration for employment without regard to race, color, religion, sex, national origin, disability, or veteran status, or any other basis protected by law.
Due to a high volume of candidates, Invisible may use automated decision-maker technologies to filter candidates based on response to our application questions and other provided information. Our use of automated decision-making enables us to be efficient by providing a manageable list of possible candidates that meet our mandatory hiring criteria. If you object to our use of automated decision-making please contact us.
About this listing
LAYIQ is an independent job-discovery service. This listing does not imply a partnership with or endorsement by the employer. Review the original posting for current details and availability.
