Bison Trailblazers | Dr. Marcus Edwards
by Anna De Cheke Qualls
Howard PhD alum Dr. Marcus Edwards is one to watch in the AI space. Co-founder of DARE Labs, he has positioned himself as a leader in cutting-edge AI technologies. His company aims to solve complex security and intelligence challenges in order to revolutionize the way organizations approach cybersecurity and threat intelligence.
A native of D.C.'s Columbia Heights, Edwards grew up playing basketball and tinkering with computers. After attending the University of Virginia, he came to Howard for his doctoral work. Living by the motto 'you only live once' has served Edwards well. Just saying 'yes' to various opportunities and people proved to be life-changing for him even in graduate school.
"It's late, I'm at the Center of Excellence during Howard's Homecoming, surrounded by empty pizza boxes, the faint scent of desperation, and the distinct aroma of grad students nine hours past their last shower. My research is hitting a wall, and I'm pretty sure my brain is about to stage a walkout. Suddenly, the door opens, and in walks Ms. [Antoinette T.] Griffin, a beacon of hope in a sea of stale coffee and existential dread. She practically herds us toward a networking event with a Netflix executive who's pitching a mentorship program. And guess who ends up being a pilot member? Yours truly. Only at Howard can you go from a code-induced coma to a career-launching opportunity faster than you can say 'free pizza’," recalls Edwards.
It was also a similar chance encounter with Dr. Danda Rawat on campus, that propelled Edwards to pursue his line of research.
"We discussed the ease with which AI can be manipulated, a concept that ignited my passion for adversarial machine learning. This field became the focus of my doctoral studies, where my research on sensor fusion and multimodal data analysis revealed a market opportunity. Thanks to Dr. Rawat's guidance and Howard's supportive entrepreneurial ecosystem, I founded DARE Labs shortly after completing my PhD, turning my research into a real-world venture working on helping AI learn in a robust way," says Edwards. "He not only steered my research but also instilled in me the crucial importance of focusing on real problems that the industry cared about."
Once in the residency program at Netflix, Edwards seized the moment. In his first year, Senior Machine Learning Engineer, Jayadev Chandrasekhar, sparked his curiosity in graph machine learning. His mentorship and insights laid the groundwork for what would become DARE Labs' core product - knowledge graphs capable of learning and adapting. In his second year, he worked with Dr. Natali Ruchansky, Manager of Machine Learning. Her encouragement to take risks helped spark Edwards' startup curiosity.
After an internship at Facebook and then graduation, Edwards didn't wait long to do his own thing. He founded DARE Labs with Demar Rogers in 2024. Its work is an extension and expansion of his doctoral research on vulnerability of AI models to data misrepresentation.
"The startup ecosystem felt like embracing controlled chaos. It demanded a deep dive into every facet of the business, from marketing and deal-making to the technical aspects. Imagine wearing ten different hats at any given moment! Unlike traditional companies where roles are narrowly defined, a startup thrives on fluidity. Networking became an unexpected adventure. I found myself connecting in the most ordinary settings: hackathons, casual gatherings, even mundane errands. My key takeaway? Cultivate intellectual curiosity. You never know who you'll encounter and what opportunities might arise," says Edwards.
His entrepreneurial journey began with identifying a real-world problem during his research and recognizing the potential for a commercial solution.
"The guiding principle? Build something you would use yourself. When it came to building a team, I tapped into my existing network. Working with people you trust is invaluable. As for the time commitment, be prepared to invest your whole self. Entrepreneurship becomes your life, shaping your schedule and priorities," advises Edwards.
Even though the company isn't doing direct research at the moment, Edwards believes that the work has broader impact. It has significant implications for ensuring truer data representations for AI.
"The ease with which data can be manipulated is a growing concern, especially with the increasing automation of critical processes by AI. This even poses a national security risk. Real-world data is inherently messy, and we're developing tools to help AI models navigate and learn from this complexity," says Edwards.
Beyond DARE Labs' public impact, there is also Edwards' positionality as an HBCU Computer Science PhD. He is part of a small and growing group of Black tech entrepreneurs. And he wants to increase representation and capacity.
"We can grow computer science departments at HBCUs by investing heavily in our computing infrastructure. This increased research capacity is essential for attracting and retaining the PhD students we need to build thriving programs," says Edwards.
According to Edwards, there are two areas where HBCUs can really lead today’s research landscape: compute-constrained research and culturally-relevant technology.
"On the compute side, while we need to build up our compute capacity, we can choose research that can is compute-constrained. Even if the next Large Language Model (LLM) that everyone is raving about doesn’t come from an HBCU, we can still make massive impact by focusing on problems where data quality and interest are the biggest blockers. There are tons of impactful spaces like small consumer robotics, simulation with synthetic data, or targeted genetics research, where smart algorithms and focused experimentation go a long way. With the right partnerships whether through university collaborations or cloud credits; we can punch above our weight and move fast without massive infrastructure," observes Edwards.
Edwards also believes that there are huge opportunities in applying AI to areas deeply connected to the diverse and historially under-explored communities.
"Things like medical disparities, hair care science, fashion technology, food systems and yield research, urban renewal are all ripe for innovation. These fields haven’t seen the same investment, not because they’re low value, but because traditional institutions just haven’t prioritized them. AI can be seen as the great equalizer in a lot of these subjects," adds Edwards.
Long-term, Edwards hopes to enable active learning for edge devices - hardware that translates, sorts and securely transfers data between a local network at the cloud (or other sources). He believes that AI that grows with new knowledge of the physical world just like people do.
His parting advice is, "go write your story, make it something you would read in terms of your career, and work backwards from there."