The first version of Wirestock was built for people with thousands of photographs and no appetite for uploading them one marketplace at a time. Each image arrived with a small bureaucracy: a caption, a list of keywords, a release when people appeared in the frame, and another submission form at another agency. Mikayel Khachatryan and three friends saw the pile of repetitive work and proposed a single door. A creator would upload once. Wirestock would handle the rest.
It was a practical idea born from both sides of the transaction. The founders had used plenty of stock imagery in earlier companies. One of those businesses was a social dating app that needed promotional photographs of happy couples. In 2018, Khachatryan began speaking with the people who supplied those marketplaces. The buyers liked the convenience. The photographers described a process that consumed time better spent shooting.
Khachatryan was an unusual fit for the visual-content trade, at least on paper. His early interests were mathematics and statistics. He moved into finance, working around trading and portfolio management, then found the field too removed from the kind of making he wanted to do. “It took me a year of working in finance to realize it wasn't for me,” he later said. A friend nudged him toward technology. By 2017, he was working on consulting and product monetization at PicsArt. Two years later, Wirestock formally began.
“Wirestock is all about making life easier for creators.”Mikayel Khachatryan, 2021
A useful mistake in the first customer profile
The team initially imagined a specialist tool for experienced stock photographers with enormous back catalogs. The first product reflected that assumption. Then Khachatryan ran workshops and watched who responded. One participant had tried stock sites without becoming a committed contributor. That person loved the simpler workflow. The reaction sent the founders toward a much larger audience: people who made commercially useful images but did not identify as professional stock photographers.
The switch reveals something about Khachatryan's operating style. He talks about quantitative subjects with ease, but the important data point was a person's relief. The initial customer model was wrong. The observed behavior was useful. Wirestock followed the behavior.
Growth created a harder problem. A friendly upload button is easy to understand. Behind it sits a factory of judgment. Images have to be captioned and keyworded. Releases need to be attached. Agencies use different formats and rules. Content arrives at volumes that make a manual spreadsheet collapse. Khachatryan has described those internal capabilities as the most technically stressful part of Wirestock's early years. The company built systems and trained review teams to process the queue.
The infrastructure hiding inside the chore
By 2020, the platform had accepted its millionth image. In 2022, Wirestock raised a $2.3 million seed round and struck a distribution agreement with Getty Images and iStock. Roughly 100,000 contributors had uploaded about three million pieces of content by the time that partnership was announced. The little administrative shortcut had become a network connecting creators with some of the industry's established outlets.
Khachatryan's public vision remained consistent: creators should spend more time creating, while software and service teams absorb the repetitive business work. He imagined more sales channels, brand projects, and tools for discovery. He also argued that demand was moving away from staged templates and toward fresh, natural scenes with real people. The observation matters because it foreshadowed the next version of the business. Useful data is not merely abundant. It is specific.
Wirestock did not just collect files. It learned how to turn loosely organized creative work into licensed, labeled, deliverable inventory.
Generative AI changed who wanted that inventory. As image models became mainstream, Wirestock received growing volumes of AI-generated art from its own users. At the same time, model developers needed visual material for training and evaluation. In 2023, the company pivoted toward supplying datasets of images, video, design assets, gaming material, and 3D content to AI labs.
This was a real operational turn, not a new label placed on the old marketplace. Off-the-shelf licensing gave way to custom requests. Teams needed to annotate and label content in greater detail. Creative contributors began completing briefs built around what a model needed to learn. Wirestock retrained parts of its workforce and eventually retired major pieces of its legacy third-party stock-distribution product.
A market for consent, context, and craft
The AI-data business rests on an argument that is simultaneously moral and commercial. Model builders need material they can use with confidence. Creators want to know how their work will be used and what they will earn. Wirestock offers rights-cleared data, contributor consent, review, and compensation as a single package. During the pivot, Khachatryan said creators were told about the change and could opt out of the data-supply business.
The platform's current projects reach beyond photographers. Its network includes videographers, illustrators, graphic and motion designers, 3D artists, filmmakers, and musicians. Some work is selected from licensed libraries. More of the business now revolves around material created to specification: a requested scene, a particular sequence of actions, a set of variations, or expert evaluation of model output.
In May 2026, Wirestock announced a $23 million Series A led by Nava Ventures, with SBVP, Formula VC, I2BF Global Ventures, and others participating. Khachatryan said the company was supplying six of the largest foundation-model makers, though he did not name them. He also put Wirestock's annual revenue run rate at $40 million and cumulative contributor payments at $15 million.
Those figures describe a sharp expansion from the stock-distribution years. They also expose the tension Wirestock has chosen to inhabit. AI systems may increase demand for human-made data while also competing with the people who make images, music, and design. Khachatryan's answer is to treat creators as suppliers with agency rather than as an unpriced archive. Whether that bargain holds depends on the details: clear briefs, informed permission, useful work, reliable review, and payments large enough to keep participation worthwhile.
“We kept building anyway.”Mikayel Khachatryan, reflecting on Wirestock's early years
The pivot kept the original customer
Plenty of pivots reveal that a company was attached to its product, not its customer. Wirestock's is more interesting because the product changed while the customer logic survived. The 2019 promise was relief from administrative work and access to more buyers. The 2026 promise is access to paid projects, with Wirestock handling the technical specifications, curation, licensing, and delivery required by AI companies.
Khachatryan has never presented himself as a photographer's romantic. His language is about systems, discovery, channels, and income. Yet the creative motive is personal. He left finance because it did not fit his temperament. He has described his co-founders as friends and Wirestock as part of their lives, work folded into the relationships rather than cordoned off from them. The company began because they enjoyed building products for creators.
That combination gives his story its shape. The statistics student notices scale. The former finance worker notices flows of value. The product operator notices friction. The person who found finance insufficiently creative notices what gets lost when administration, distribution, or a platform captures too much attention.
Wirestock now occupies a disputed piece of the AI economy. Its job is to make human creativity legible to machines and valuable to the humans who supplied it. The work is less cinematic than the model at the end of the pipeline. It is releases, labels, edge cases, review queues, and payment records. Those details are precisely where a fairer market either becomes real or remains a slogan.
Authenticity became a specification
Years before the AI pivot, Khachatryan was already telling stock contributors that buyers preferred natural images of real people doing real things. He encouraged work that felt current and specific, including pictures made on phones. The advice sounded like a market tip. It later became an infrastructure problem. If a customer needs a precise action, object, setting, or cultural context, a generic archive may not contain the right example. A commissioned network can ask for it.
That shift changes the creator's role. Instead of waiting for an unknown buyer to discover a file, a photographer or designer can respond to a defined need. The model builder gets material shaped for a task. The platform can document where it came from, how it was reviewed, and what permission travels with it. The creator gets a clearer line between effort and payment. None of those steps removes the imbalance between a large technology buyer and an individual contributor, but each makes the transaction easier to inspect.
Khachatryan's earliest insight was that creators should not have to repeat the same clerical task to reach each new buyer. His latest bet carries that insight into a more consequential market. The machine may be new. The founder is still working on the messy middle.