Shipping ought to be the boring part. A scientist has done the experiment, a company has made the kit, and somebody puts it in a box. Siddarth Selvaraj discovered that the box deserved more attention. After Arima Genomics sent a beta version of its laboratory kit to prospective customers, he had a wonderfully plain verdict: ‘Shipping, we realized, was not a simple problem.’ The genome had supplied one set of difficulties. Getting a usable product into another scientist’s hands supplied another.
That small admission makes a useful entrance to his career. Selvaraj, who also goes by Sid, built Arima around research into DNA’s three-dimensional organization. He spent a decade as its founding CEO, then moved into an operating role when Tom Willis became chief executive in June 2025. The work has required him to cross several borders: between computing and biology, between a published result and a repeatable workflow, and between explaining an idea and persuading someone to pay for it.
A method that works once must survive repetition before it can travel.From the laboratory to the customer
The genome had a seating plan
At the University of California San Diego, Selvaraj studied bioinformatics. His doctoral thesis examined three-dimensional genome organization and gene regulation in mammalian cells; his advisers were Bing Ren and Vineet Bafna. The subject asks a deceptively awkward question. If DNA contains information in its sequence, what additional information sits in the way that sequence is arranged? A string can be long on paper and folded into close company inside a cell. Distance along the string and distance in space describe different relationships.
In 2013, Selvaraj, Jesse Dixon, Vikas Bansal and Ren published work in Nature Biotechnology on reconstructing whole-genome haplotypes using proximity ligation and shotgun sequencing. A haplotype groups variants belonging to the same chromosome copy. The team’s approach used physical relationships within the nucleus to help work out which variants belonged together. Spatial organization became evidence that could assist the reconstruction of sequence. For a reader accustomed to thinking of DNA as four letters in an exceptionally long sentence, this introduces a second grammar.
The collaboration matters as much as the concept. This was work shared among researchers with experimental and computational expertise. Selvaraj’s later company emerged from that setting, where a result depends on the meeting of an assay, a data set and an analysis. His scientific record also includes a 2013 paper on Bayesian inference of chromosome organization. The recurring interest is tangible: use measurements and computation to understand how a genome occupies space.
An idea learns to leave the laboratory
Selvaraj dates Arima’s founding to 2015 in his own account of the decade. He described it as an attempt to commercialize what he had learned during his PhD. That is a concise origin story, but it leaves an enormous practical question sitting on the bench: what, exactly, should the customer receive? A scientific finding does not come with a distribution model attached. A company has to decide how much of the experiment to package, how much expertise to assume and how the result will fit into somebody else’s day.
Arima entered the I-Corps entrepreneurial training program in 2017. Across its eight-week course, the team examined where its offering belonged in a larger research workflow and conducted customer outreach. Selvaraj credited that process with helping establish the kit-based distribution model. The company also used early grant support for scientific validation, then later support to improve protocols, source materials and develop manufacturing capabilities. These were separate jobs, requiring separate answers. A method that works once must survive repetition before it can travel.
The commercial ambition appears in a 2017 grant proposal with almost comic specificity. Arima planned to improve sourcing, reduce cost, design for manufacture and establish performance under real shipping and storage conditions. It wanted a product compatible with existing short-read sequencing workflows, without requiring customers to acquire unusual equipment or a new collection of specialist skills. The proposal makes the operator’s task visible. Each extra demand on the buyer is another opportunity for an otherwise promising method to remain an interesting paper.
Beta customers supplied the sort of evidence no internal meeting can produce. They received the kit, used it and reported what they liked and what needed changing. Commercial kit sales began in 2018. The shipping lesson belongs in that sequence because it exposes the difference between controlling an experiment and controlling a customer experience. A package, its contents and its instructions all arrive together. The recipient does not have the founder standing beside the bench to explain the intention.
A different audience for the same idea
By October 2020, the Wellcome Sanger Institute had evaluated and implemented Arima’s high-coverage Hi-C technology for projects including Darwin Tree of Life. The criterion was agreeably practical: the kits needed to work across tissue types from diverse vertebrate and invertebrate species. In this setting, success meant helping researchers assemble genomes from material that refuses to behave like a perfectly tidy laboratory example. A company built around spatial information was finding its place in the reconstruction of linear genomes.
Early users also tested limited sample amounts and challenging storage conditions. Feedback helped Arima optimize the offering. These details give the customer-discovery story a measurable consequence: researchers were testing the technology against the constraints of their own work. The kit had to meet those constraints to become useful. Laboratories have a healthy talent for spoiling an elegant sales pitch with an inconvenient sample.
In September 2021, Arima announced a collaboration with the European Reference Genome Atlas, supplying Hi-C kits and reagents for reference-genome assembly. The company had also established a distribution partnership in Japan and worked with the Vertebrate Genomes Project. These are company achievements, built by a team, during Selvaraj’s tenure as CEO. Their significance for his story is the widening audience. Research begun in a university environment had become something other institutions could purchase and incorporate into their own programs.
The recruitment conversation that worked
Building that business also meant getting people to join it. On Selvaraj’s LinkedIn profile, Derek Reid recounts initially choosing a job elsewhere, then returning for a second interview. Selvaraj discussed the technology and the growth the team could pursue. Reid says that conversation changed his decision. It is a more personal measure of persuasion than a financing announcement: someone who expected to leave the room unconvinced decided to come to work.
Another colleague, Jay Nakamura Clark, describes an effort to create an inclusive, family-like company culture, alongside technical knowledge and product strategy. These are colleagues’ assessments, and they show how he appeared to people working with him. A founder can explain an assay fluently and still struggle to explain a workplace. The recommendations suggest that Selvaraj devoted attention to both. The practical connection is easy to understand: a product takes people to build, improve, support and sell it.
His own public profile allows a small joke into the credentials. Alongside English and Tamil, he lists Martian as a language. No interplanetary certification is required to appreciate the entry. It puts a little mischief beside the doctoral qualification. In his end-of-decade reflection, he used the hashtags #benice and #celebratescience. The pairing is a useful glimpse of the register he chose for himself: ambitious about the work, with room for warmth.
Advice earned by rewriting
In a 2024 public conversation about commercialization, Selvaraj explained that his advice came from writing and rewriting grant applications. His test for a customer need was action: would somebody do something about the problem? He urged founders to talk repeatedly with customers, document the conversations and seek feedback from investors and other CEOs. He also cautioned against expecting an investor to write a check immediately. Trust, like a workable protocol, takes time to develop.
He returned to that theme in a later discussion, treating investors as another kind of customer whose feedback could clarify a business. He described selling into the research market first as a way to improve the product, build credibility and earn revenue while the company developed further. The advice has the texture of a working founder’s experience. It concerns the order in which difficult things can be attempted, and the money and evidence needed to keep attempting them.
There is something pleasing about a researcher recommending more conversations and better documentation. A commercial plan becomes a proposition to test. Each interview can change the assumptions; repeated feedback can reveal a pattern. His account does not suggest that the answers arrive in one inspired afternoon. The useful part is the willingness to revise. That same willingness appears earlier in the beta kits and later in the response to researchers handling different samples.
The founder changes seats
On June 17, 2025, Arima announced a $22 million Series C led by Illumina Ventures and the appointment of Tom Willis as CEO. Selvaraj would continue as president and chief operating officer. The current leadership page lists him as founder and COO, and includes him on the board of directors. After ten years in the chief executive’s seat, he remained inside the company he had helped build. Its leadership arrangement changed while his operating involvement continued.
A financing round can make a career look like a series of neatly marked milestones. Selvaraj’s account supplies the less tidy intervals: the customer calls, the applications rewritten, the product sent out and adjusted. Those details make the milestones intelligible. His original problem involved preserving relationships within DNA. His business required attention to relationships among researchers, colleagues, investors and the people receiving a box. The box was always part of the experiment.