Democratizing
Vision AI
Making advanced Computer Vision accessible to everyone—not just big tech companies with massive budgets.
Why Dhakma Exists
"For years, I watched advanced Vision AI remain locked behind the walls of big tech companies and elite research labs. Only teams with million-dollar budgets could access the expertise needed to build sophisticated computer vision systems."
"That never sat right with me."
"I started Dhakma with a simple belief: every company—whether it's a startup with a breakthrough idea or an established business expanding into AI—deserves access to the same caliber of Computer Vision expertise that tech giants have. At a fraction of the cost."
"Today, we're helping companies around the world harness the power of Vision AI and audio-visual applications—achieving what larger companies can, but faster and more affordably. That's democratization in action."
Who We Are
Dhakma was founded by Dr. Tharindu "Mackie" Mathew, who brings a unique combination of deep academic research (PhD from Purdue University) and enterprise-scale engineering (Ex-Microsoft). What started as a mission to democratize Vision AI has grown into a global team of specialists.
We're headquartered in West Lafayette, Indiana—right next to Purdue University, giving us direct access to cutting-edge research and academic partnerships. Our US-based location makes us ideal for enterprises, startups, and funded research projects. But our reach is global: with team members across North America, Europe, and Asia, we provide expertise and coverage that spans time zones and borders.
Research Implementation, Not Just Tools
We don't wrap APIs—we implement papers. From CVPR to ICCV, we bring cutting-edge research to production before it's productized elsewhere.
US-Led, Globally Distributed
Our model combines US-based leadership and compliance with global talent, allowing us to deliver world-class Vision AI expertise at prices that make sense for growing companies—not just tech giants.
Custom Algorithms, Not Cookie-Cutter Solutions
While platforms like Roboflow provide excellent tools for standard use cases, we go deeper—implementing state-of-the-art research papers and building custom architectures when your problem demands it.
Our Mission
We believe the next generation of breakthrough Vision AI applications won't come from big tech—they'll come from innovative startups and forward-thinking companies solving real problems across industries.
Our job is to give them the tools to compete.
By translating cutting-edge research into production-ready systems. By making advanced techniques accessible within weeks, not years. By pricing our expertise based on value delivered, not hours billed. This is how we democratize Vision AI.
How We Work
Research Meets Production
We read the latest computer vision papers (CVPR, ICCV, ECCV) and implement them from scratch—adapting novel architectures to your specific problem. No waiting for API providers to add features. No limitations of pre-built tools. Every prototype we build is designed for scale.
Milestone-Based Pricing
You pay for results, not hours. We break projects into clear milestones with deliverables you can verify before paying. This keeps costs predictable and aligns our success with yours.
Your Extended R&D Team
We become your Computer Vision R&D team—available on-demand without the overhead of full-time hires. Whether you need us for one project or ongoing development, we scale with your needs.
Vision AI & Audio-Visual Applications
From perception systems to visual understanding, from real-time processing to large-scale analysis—if it involves computer vision or audio-visual AI, we build it.
We don't limit ourselves to specific industries. Whether you're in healthcare, manufacturing, agriculture, entertainment, security, autonomous systems, or an entirely new domain—we bring the same research-grade expertise and production engineering to solve your vision challenges.
Ready to Access
World-Class Vision AI?
You don't need a big tech budget to build advanced Computer Vision systems. Let's talk about your project.