Company Overview:

We are building Protege to solve the biggest unmet need in AI — getting access to the right training data. The process today is time intensive, incredibly expensive, and often ends in failure. The Protege platform facilitates the secure, efficient, and privacy-centric exchange of AI training data.

Solving AI’s data problem is a generational opportunity. We’re backed by world-class investors and already powering partnerships with some of the most ambitious teams in AI. The company that succeeds will be one of the largest in AI — and in tech.

We’re a lean, fast-moving, high-trust team of builders who are obsessed with velocity and impact. Our culture is built for people who thrive on ambiguity, own outcomes, and want to shape the future of data and AI.

About the Role

Protege is hiring an FDE to support our audio vertical. This is an engineering role with two key mandates: own the technical success for all audio customers, and build the reusable pipes that enable future scale. You will partner closely with the GM of the vertical, product and platform engineering, Data Lab, and commercial teams to navigate external requirements and develop solutions to execute on those requirements. Audio is a rapidly developing and evolving area at Protege — this role will be instrumental in defining how the function operates.

This role is ideal for engineers who prioritize both constantly learning through solving difficult problems and interfacing and directly owning customer outcomes. You'll be responsible for managing customer requests, newly ingested partner datasets and architecture decisions all at the same time, often across multiple deals. The pace is fast, the ambiguity is real, and when deals are live, availability outside standard hours is part of the job. If this type of environment and ownership excites you, there could be a strong fit.

What You'll Do

Build the Technical Foundation for Audio

• Develop and own the solutions that turn raw audio, with often unstructured or non-existent metadata, into organized and ready-to-sell datasets.

• Build prototypes and phase them into reliable production systems that others can understand and operate.

• Rapidly characterize large and unfamiliar audio datasets and determine whether they can support a customer's use case.

Own Customer Engagements End to End

• Work with the GM of the audio vertical, core technical teams, and commercial stakeholders from feasibility through post-delivery support.

• Translate a customer's model-development goals into an executable technical plan with clear acceptance criteria.

• Own the implementation and operation of the engagements you lead.

Turn Deal Work Into Product Leverage

• Work in tandem with the product and core engineering teams to identify areas of growth for the platform based on customer requests and learnings from the front lines.

• Identify recurring patterns that should become shared cross-vertical platform capabilities and contribute directly to designing and building them.

• Build the audio vertical's technical playbooks and quality standards so the function scales with every new request.

What Success Looks Like

30 Days: Learn and Ship

• Learn Protege's platform, audio data landscape, active customer and partner portfolio, and current processing and delivery systems.

• Pair with an engineer to work through an existing audio deal to understand what a delivery looks like at Protege.

• Map the largest gaps in the audio vertical's tooling and operating model and propose a prioritized plan for addressing them.

60 Days: Own End-to-End

• Operate and run an active audio deal as the primary FDE.

• Build or meaningfully extend a tool or workflow that spawns from an active customer request.

• Establish a communication cycle with other FDEs and product to surface learnings from deals to identify areas of generalization.

90 Days: Operate Independently

• Serve as the default technical owner across the audio vertical's active portfolio, including owning multiple active deals.

• Own end-to-end architecture and delivery for audio customers, including post-delivery support and iteration.

• Establish the first version of the audio FDE playbook and reusable toolkit.

• Maintain a concrete roadmap of platform investments with the goal of increasing the vertical's delivery capacity.

What You Bring

Must Haves

• 3+ years of experience as an engineer, including meaningful exposure to customers or external technical stakeholders.

• Experience working directly with media data, bonus points for direct work with audio data.

• Experience building and operating systems that process, analyze, or deliver data at scale.

• Customer-facing ability, including translating ambiguous requirements, communicating trade-offs, and building trust with technical stakeholders.

• Demonstrated end-to-end ownership, from initial problem definition through implementation, validation, and support.

• High ambiguity tolerance and bias to action, with the judgment to know when to investigate further or push back.

• Comfort with the intensity and pace of a fast-moving environment, including multiple concurrent priorities and time-sensitive customer work.

Nice to Haves

• Experience at an early-stage company, as a founding or early engineer, or in another startup-like environment with broad ownership.

• Hands-on experience with Python.

• Product engineering experience or a strong product mindset developed in close partnership with users.

• Experience with search, vector embeddings, semantic retrieval, or ML-assisted data curation.

• Experience with our cloud and data infrastructure (AWS, Dagster, Databricks, Vercel are the key tools).

• Experience with audio processing, speech or music ML, transcription, diarization, annotation, quality measurement, or privacy and redaction workflows.

Protege Values

Pass the Loved Ones’ Test

We act with integrity and do the right thing — especially when it’s hard and no one is watching.

Always Find a Way

We are resourceful, resilient builders who solve hard problems and push through obstacles.

Go Fast and Grow Fast

Velocity matters. We move with urgency, learn quickly, and continuously improve as individuals and as a company.

Practice Kindness and Candor

We communicate directly and respectfully, building trust through honest feedback and genuine care for one another.

Deliver Together

We win as one team. Collaboration, accountability, and shared ownership drive our success.

Own the Outcome. Hone the Craft.

We take pride in our work, sweat the details, and continuously raise the bar for excellence.

About the company

Protege operates as a data marketplace that connects data holders with AI developers to enable secure, compliant exchange of training data for AI model development. Its platform streamlines data procurement by listing diverse and curated datasets, including specialized healthcare data, and facilitating data licensing deals between buyers and data owners. The acquisition of Calliope Networks adds premium video content, expanding into multimodal data to support generative AI models. Revenue likely comes from transaction fees on licensing deals or subscription access for data buyers. The goal is to provide ready access to high-quality, diverse training data while ensuring data security and regulatory compliance, helping AI teams accelerate model development.

Company Size

51-200

Company Stage

Series A

Total Funding

$65M

Headquarters

New York City, New York

Founded

2024

Get referred to Protege

See people who can refer or advise you

Simplify's Take

What believers are saying

Protege closed a $30 million January 2026 extension, reaching $65 million total funding.
DataLab won backing from a majority of Magnificent 7 companies by March 2026.
February 2026 healthcare benchmarks and lab-data deals deepen demand in regulated, high-value markets.

What critics are saying

Copyright litigation over AI training data escalates in 2026, raising indemnity and sourcing costs.
OpenAI, Anthropic, and big labs can bypass marketplaces by licensing directly from suppliers.
If licensors demand higher rev shares, Protege’s margin and network effects collapse quickly.

What makes Protege unique

Protege sells rights-protected, real-world multimodal data, not generic labeling or scraping.
DataLab, launched March 2026, turns dataset design into research, benchmarks, and methodology.
Partnerships with Dailymotion, HC1, and healthcare providers expand proprietary supply across verticals.

Help us improve and share your feedback! Did you find this helpful?

Benefits

Remote Work Options

Growth & Insights and Company News

Headcount

6 month growth

-4%

1 year growth

11%

2 year growth

26%

HPCwire

Mar 12th, 2026

Protege Launches DataLab to Turn AI Data Into a Scientific Discipline

Protege launches DataLab to turn AI data into a scientific discipline. Press play to listen to this content At BigDATAwire, we have extensively covered how progress in AI is no longer driven only by larger models or more compute. Increasingly, the real constraint is the data itself, and that bottleneck becomes harder to solve as systems grow more complex and move closer to real-world use. Protege, an AI data platform focused on delivering large-scale, real-world datasets for model training, is building on that idea with the launch of DataLab, a new research initiative intended to bring more rigor, standards, and scientific discipline to the AI data layer. DataLab is essentially a research institution intended to help AI researchers navigate the growing challenges around the science of data, including dataset quality, selection, representation, and evaluation. The DataLab team includes in-house experts and researchers. Protege claims that most of the Magnificent 7 tech companies and multiple frontier AI labs are participating in early training and evaluation data collaborations. The Magnificent 7 includes Amazon, Apple, Alphabet, Microsoft, Nvidia, Meta, and Tesla. The launch of DataLab comes at a time when AI development is increasingly shaped by data limitations. A recent Snowflake survey found strong returns from GenAI projects but also highlighted ongoing challenges around data readiness and quality. This reinforces the idea that future progress will depend as much on better data as on larger models or more compute. "We understand the three core pillars driving AI: models, chips, and data. We are convinced that with the right datasets - the third, underdeveloped pillar - you can push the entire frontier forward," said Bobby Samuels, CEO of Protege. Samuels says that the company "created DataLab to treat data as infrastructure, not exhaust." He argues that to get more capable, reliable systems, you need to have better standards, reproducibility, and real scientific discipline at the data layer. The three core areas for DataLab include scientific partnerships, building high-value datasets and data products, and leading AI data research. The work is expected to span both academic research and commercial applications, with the lab exploring new product opportunities while also publishing benchmarks and technical studies. "The strength of DataLab is its ability to integrate perspectives that are often siloed," said Engy Ziedan, Co-Founder and Chief Scientific Officer at Protege. "Advancing AI requires more than larger models or more data alone." Ziedan emphasizes that this "requires thinking at the margin, where we weigh the marginal value of a datapoint on learning and the opportunity cost of choosing the wrong dataset." He believes that his team is structured to deliver the disciplined dataset design, careful evaluation, and a deep understanding of real-world complexity required for frontier AI. One reason efforts like DataLab are getting attention now is that AI systems are moving into areas where mistakes are harder to tolerate. Training a model on internet data is one thing. Training systems that help with scientific workflows require a very different level of precision. The challenge is not just finding more data, but finding the right data, understanding how it was collected, and knowing how it affects the outcome of a model. Researchers have also started to focus more on the marginal value of data, meaning how much a single dataset or even a single datapoint changes model behavior. As models become larger, simply adding more information does not always improve results. In some cases, the wrong data can reduce performance or introduce errors that are difficult to detect. That makes dataset design a technical problem in its own right rather than just a prep step before training. Protege says DataLab is meant to work at that level, where decisions about what data to include, how to structure it, and how to measure its impact can determine whether a system performs reliably outside the lab. As AI moves further into real-world use, that layer is becoming harder to ignore. If you want to read more stories like this and stay ahead of the curve in data and AI, subscribe to BigDataWire and follow us on LinkedIn. We deliver the insights, reporting, and breakthroughs that define the next era of technology. The latest Snowflake report titled "The ROI of Gen AI and Agents" shows that GenAI... The AI fatigue that defined the late 2023 and 2024 business cycles was, in hindsight,... MariaDB is set to acquire GridGain Systems with the aim to deliver sub-millisecond data performance... In 2008, scientists did something extraordinary. For the first time, they built the full genetic... Moving large-scale data across platforms, clouds, and global regions is no longer a special project... Deloitte's latest State of the AI report shows that AI adoption continues to accelerate rapidly,...

Business Wire

Mar 11th, 2026

Protege launches DataLab to advance AI data science with backing from Magnificent 7 companies

Protege, an AI data platform, has launched DataLab, a research institution focused on advancing the science of AI training and evaluation data. At launch, a majority of the "Magnificent 7" AI companies and major frontier AI labs are collaborating with DataLab on various projects. Led by Engy Ziedan, Protege's co-founder and chief scientific officer, DataLab addresses data quality, selection and methodology challenges as AI development becomes increasingly constrained by data limitations. The institution operates across three areas: scientific partnerships with AI researchers, building high-value datasets, and conducting AI data research. DataLab has already released multimodal healthcare benchmark datasets and is working with frontier AI organisations on challenges ranging from advanced cancer diagnostics to audio de-identification. The initiative aims to bring scientific rigour to dataset design and establish reproducible methodologies for more reliable AI systems.

SuperbCrew

Jan 9th, 2026

Protege Raises $30M in Series A Funding Led By a16z

Protege raises $30M in Series A funding led by a16z. Protege successfully raised $30 million in a Series A extension round led by Andreessen Horowitz (a16z), with participation from returning investors including Footwork, CRV, Bloomberg Beta, Flex Capital, and Shaper Capital, elevating the company's total funding to $65 million since its 2024 founding. Protege operates as an AI data platform that facilitates access to trusted, real world datasets at scale. Founded in 2024 by Bobby Samuels (CEO), Travis May (Chairman, with prior experience at Datavant and LiveRamp), Richard Ho (CTO), and Engy Ziedan (Chief Scientific Officer), the company emphasizes ethical sourcing of multimodal data, including de-identified health records, medical imaging, audio recordings, and media content. It curates datasets for AI training and evaluation, partnering with data providers through licensing agreements and offering revenue sharing models. By 2025, Protege had expanded its network to hundreds of organizations and supports workflows for leading AI institutions worldwide. This latest round builds on a $25 million Series A in August 2025 led by Footwork and a $10 million seed round led by CRV. The extension underscores rapid adoption in industries facing data shortages for AI development. a16z Partner Daisy Wolf noted that Protege's approach respects data complexities while enabling modern AI use, highlighting a market shift toward responsible data unlocking. CEO Bobby Samuels emphasized the platform's role in supplying curated, AI ready data amid fragmented sources, while Chairman Travis May pointed to proprietary data as the driver for AI's next phase. Protege's $30 million Series A extension round, led by Andreessen Horowitz (a16z) and announced on January 8, 2026, marks a significant milestone in the company's trajectory, bringing its total funding to $65 million since its inception in 2024. This funding builds directly on the momentum from Protege's prior raises, including a $25 million Series A in August 2025 led by Footwork and a $10 million seed round led by CRV, reflecting investor confidence in its mission to address one of AI's most pressing challenges: access to high quality, real world data. At its core, Protege functions as a governed marketplace and data infrastructure platform that connects data holders (such as organizations in healthcare, media, audio, and motion capture) with AI developers seeking proprietary, multimodal datasets for training, fine tuning, and evaluation. Founded by a seasoned team including CEO Bobby Samuels, Chairman Travis May (former CEO of Datavant and LiveRamp), CTO Richard Ho, and Chief Scientific Officer Engy Ziedan, the company has rapidly scaled its partner network to hundreds of organizations by 2025, emphasizing ethical licensing, data curation, anonymization, and revenue sharing models that compensate providers based on usage. This approach differentiates Protege from reliance on public or synthetic data, focusing instead on real world sources that capture authentic human and system behaviors across domains like video, imaging, gaming, manufacturing, life sciences, real estate, finance, and education. The funding round arrives amid a broader industry shift where AI progress is increasingly constrained by data availability rather than compute or model architecture. Public datasets have been largely exhausted, and the internet's scrapable content has reached its limits, pushing developers toward fragmented, proprietary sources that are often inaccessible due to privacy, intellectual property, and operational hurdles. Protege addresses these by streamlining discovery, filtering, and combination of datasets with built-in compliance and transparency, effectively shortening delivery timelines from years to months in sectors like healthcare. As Travis May articulated, "Access to data is the biggest bottleneck to the advancement of AI. The next phase of AI will be driven by real world, proprietary data generated through everyday human activity." Similarly, Bobby Samuels highlighted the demand supply imbalance: "We're seeing demand for real world data grow faster than the market's ability to supply it responsibly." Investor enthusiasm, particularly from a16z, stems from Protege's proven product market fit, as evidenced by its collaborations with foundational model builders, including the majority of the Magnificent Seven tech giants. Daisy Wolf of a16z remarked, "The next era of AI will be shaped by who can responsibly unlock access to the world's most valuable data," underscoring the platform's role in navigating complex data landscapes. The capital will fuel specific initiatives: accelerating product features for data cleaning and formatting, broadening coverage into new verticals, enhancing partnerships, and expanding the team across roles like data scientists, engineers, and operations. In the competitive landscape, Protege stands out by concentrating on data aggregation and ethical exchange, unlike broader AI tooling providers. Key competitors include Scale AI, which offers data labeling and annotation services; Snorkel AI, focused on programmatic data labeling for machine learning; and Labelbox, which provides tools for data labeling and management. These players address adjacent needs but lack Protege's emphasis on proprietary, real world data marketplaces with revenue sharing and compliance layers. Looking ahead, this funding positions Protege to influence AI's evolution, potentially shaping standards for data valuation, licensing, and ethical AI development, as seen in its participation in events like CES 2026 panels on AI copyright and data valuation. Reactions from the tech community, including congratulatory notes on LinkedIn and X, indicate strong support for Protege's vision amid predictions that media licensing for AI could reach hundreds of millions in revenue by 2026 alone. Funding History Table | Round | Amount Raised | Lead Investor | Date | Participating Investors | Cumulative Total | | Seed | $10 million | CRV | 2024 (exact date not specified) | Not detailed | $10 million | | Series A | $25 million | Footwork | August 2025 | CRV, Bloomberg Beta, Flex Capital, Shaper Capital | $35 million | | Series A Extension | $30 million | Andreessen Horowitz (a16z) | January 8, 2026 | Footwork, CRV, Bloomberg Beta, Flex Capital, Shaper Capital | $65 million | | Investor | Role in Latest Round | Notable Background/Contributions | | Andreessen Horowitz (a16z) | Lead | Focus on AI infrastructure; Partner Daisy Wolf emphasizes data's role in AI's future. | | Footwork | Returning Participant | Led previous Series A; Co-founder Nikhil Basu Trivedi supports data focused ventures. | | CRV | Returning Participant | Led seed round; Partner Saar Gur invests in enterprise tech and AI. | | Bloomberg Beta | Returning Participant | Early stage AI and data investments aligned with media and tech. | | Flex Capital | Returning Participant | Focus on scalable tech platforms. | | Shaper Capital | Returning Participant | Supports innovative data and AI startups. | | Company | Primary Focus | Key Differentiation from Protege | Funding/Scale Highlights | | Scale AI | Data labeling and annotation for AI models | Broader tooling including human in the loop labeling; less emphasis on proprietary data marketplaces. Raised over $1 billion; serves major AI firms. | | Snorkel AI | Programmatic data labeling and management | Focuses on weak supervision and custom labeling pipelines; not centered on real world data aggregation. $135 million in funding; enterprise oriented. | | Labelbox | Data labeling platform with collaboration tools | Emphasizes workflow for labeling teams; lacks revenue sharing for data providers. $188 million raised; integrates with ML frameworks. | This round not only validates Protege's model but also signals a maturing AI data ecosystem where ethical, scalable access could unlock trillions in economic value, though ongoing debates around data privacy and ownership will shape its long term impact.

SiliconANGLE Media

Jan 8th, 2026

Protege raises $30M to grow governed marketplace for AI training data

Protege raises $30M to grow governed marketplace for AI training data - SiliconANGLE

TechStartups

Aug 13th, 2025

Top Startup and Tech Funding News - August 13, 2025

August Health locked in $29 million to advance its AI-powered senior care platform, while AI data platform Protege raised $25 million in Series A funding.

Protege

Forward Deployed Engineer - Audio

About the job