Leading a global open source project with hundreds of contributors across borders, cultures, platforms, and organizations is a prospect fraught with challenges.
Deven Panchal is well aware of those challenges, having led two such projects for AT&T Labs:
- Acumos, the world’s first open source AI marketplace
- ONAP, which allowed developers to embed AI into the design and deployment of network, telecom, and security services
To manage—if not avoid—the myriad complications entailed by such endeavors depends, Panchal says, on multiple factors, from a shared vision to radical transparency to interaction-tracking mechanisms.
It also requires trust, both that people will do the right thing and that inevitable conflicts will be respectfully resolved.
All of these factors require skilled leaders at the helm; the best, he says, have “worn many hats.” Panchal’s varied industry experiences, leadership skills, and innovative work certainly qualify him on all fronts. They also earned him a spot among Computing’s Top 30 Early Career Professionals for 2024.
In the following Q&A, Panchal describes
- Details of the ground-breaking ONAP project, including its impact on people and networks across the world
- The pioneering concepts he developed for use in the global Acumos AI marketplace, including Reusable MLOps to allow organizations to use standardized AI model pipelines
- How staying hands-on and engaged with actual technology can help leaders sort through the AI hype and make decisions based on fundamental principles and emerging capabilities
- The role of AI in organizations, as well as the need to have responsible processes in place to realize its tremendous potential
Your career journey is quite remarkable—from doing pioneering work in India to leading significant AI and network software initiatives at AT&T, a global giant. Can you walk us through some pivotal moments that shaped the journey?
Absolutely. My journey started with a deep fascination for how computers and communications could connect us. Early on, at India’s Society for Applied Microwave Electronics Engineering and Research (SAMEER)—a semi-government R&D lab—I was part of the team that built India’s first indigenous and most affordable 6MeV LINAC and MLC for cancer radiotherapy.
This experience, working at the intersection of software, core research, and go-to-market productization for social good was incredibly special. It taught me the tangible impact technology can have. At AT&T Labs in the U.S., I led global open source projects like Acumos AI and ONAP. These were massive, cross-company, cross-geography collaborations.
For instance, Acumos, which I co-created, became the world’s first open source AI marketplace, democratizing AI and being adopted by initiatives like the UN’s AI4Action and Europe’s AI On-Demand Ecosystem. Similarly, ONAP, where I led key components like DCAEMOD—which allowed the embedding of AI in the design and deployment of network, telecom, and security services—has revolutionized network automation for telcos worldwide, impacting billions of subscribers globally. These experiences, operating in diverse environments and leading global teams, have given me great breadth as well as depth in both technology and business. I am very grateful for these experiences.
You’re credited with pioneering concepts like Reusable MLOps, N-way AI Federation, and Sellable AI within Acumos. Can you elaborate on their significance and how they’ve shaped the AI landscape?
Certainly. When we envisioned Acumos, the goal was to make AI models as easy to share, combine, and deploy as apps in an app store. Reusable MLOps was a foundational concept I introduced, focusing on creating standardized, hot-swappable AI model pipelines. This means organizations don’t have to reinvent the wheel; they can leverage pre-built, tested components, drastically reducing development time and cost. N-way AI Federation addressed the challenge of proprietary data and models by enabling different entities to collaborate, share, and even license AI models securely and ethically, fostering a richer ecosystem. And the idea of Sellable AI built on this, creating a marketplace for AI assets. These concepts were implemented and have been adopted globally, making AI more accessible, collaborative, and more impactful.
Leading such global, collaborative projects must come with unique challenges. How do you ensure alignment and project delivery across diverse—and often geographically dispersed—teams and organizations?
Leading these global efforts—like the Acumos and ONAP initiatives, which involved hundreds of contributors from multiple tech giants and academic institutions worldwide—requires several things:
- Establishing a clear, shared vision
- Fostering a culture of open communication and radical transparency
- Having mechanisms and metrics to track various human and non-human interactions (which open source has become excellent at)
- Recognizing that conflicts will arise but can be negotiated and respectfully resolved
- Trusting the teams and collaborators to do what is right
I also believe that leaders must have the skills to understand every aspect of the project and details when required. This can only be achieved through the experience of having worn multiple hats in your career. I have myself had multiple technical and managerial experiences in my career, from product owner/manager to the lead engineer to architect to scrum master and so on. I’ve always believed in leading by example.
You’ve received numerous prestigious awards, including multiple AI Innovator of the Year awards, AI100 Generative AI Leader award, and multiple Best Paper awards at international conferences. How do you view these accolades in the context of your work?
I’m incredibly grateful for these recognitions, but I truly see these awards as a by-product of impactful, collaborative work. These are hard earned, but the real reward is seeing a project like ONAP adopted by telcos serving over 70% of global mobile subscribers, or Acumos enabling new AI solutions that tackle real-world problems. The Best Paper or Best Presenter awards are due to the novel concepts or solutions developed within these global projects and presented at international conferences in places like Australia, Indonesia, Egypt, Vietnam, and Morocco. They reflect the innovation happening within the teams I’ve been privileged to lead and contribute to. The true measure of success is the tangible positive change we create.
The AI field is currently buzzing with activity—some might say “noise.” As a leader who’s deeply hands-on, how do you cut through this noise to identify truly impactful AI strategies?
I want to state outright that you cannot ignore AI, and it will be hugely transformative. But you’re right, there’s a lot of noise in AI, and it’s easy to get caught up in the hype. My approach is to stay grounded in fundamental principles while being aware of emerging capabilities. Being hands-on is non-negotiable for me, even in senior roles. One of the biggest challenges I see in the industry today is leaders who become too detached from the actual technology. As a CTO or chief AI officer, you need to understand the “how,” as well as the “what” and the “why.” I write open source software (I recently created simpliflow—a gen AI agentic framework), participate in architecture designs, contribute to strategic technical discussions, mentor AI and software engineers, publish research and implementation papers, am a regular at many industry events, and participate in international conferences, which allows me to present globally and connect with industry and academia.
This allows me to swiftly and decisively separate what is worth paying attention to versus what’s not. This, coupled with the knowledge of the business and the financial objectives, create a holistic view that can help you make intelligent decisions wherever you operate. You have to keep learning.
How has working with international teams and participating in global conferences influenced your approach to problem-solving and innovation?
Leading and collaborating with international teams on projects like Acumos and ONAP—and presenting my research at global conferences—has exposed me to a rich tapestry of perspectives, methodologies, and cultural approaches to problem-solving. This global interaction breaks down echo chambers and challenges assumptions. It has taught me the importance of clear communication and cultural sensitivity. These perspectives are invaluable as we build products for a global consumer base.
You are actively involved in various IEEE Standards committees and open source communities. How has this involvement influenced your work and professional growth?
It allows me to shape the future of ideas and also exposes me to them before they hit the industry, which helps me to bring those insights early on to my work. Like open source, the “global default” reduces re-work and makes solutions scalable by design.
You have been involved in filing numerous patents. Can you share the importance of patenting in your field?
I’ve filed several U.S. patent applications for concepts developed within projects like Acumos and ONAP, often as the first inventor. At a higher level, I believe that innovation is the synonym of survival. Patents do one job of converting these innovations into strategic moat. But innovating pays back in multiple ways. So, I believe anyone can and should be empowered to innovate.
Very generally, what will we see of AI and Generative AI in the near future?
Most definitely: more powerful models and their integration and interaction with software systems, hardware systems, and our physical world.
Looking ahead, what’s your vision for the role of AI in large enterprises?
There is no doubt that current technology AI allows us to operate leaner and smarter than ever before. So, the onus is on us to have processes to responsibly realize this potential whether it is in an enterprise setting or elsewhere. Our focus should then be to do our job better and be better versions of ourselves, no? In an enterprise setting, this could mean enhanced customer experiences, better reliability, new products that “wow” people, etc. Many enterprises globally are building this AI-first culture and, as a part of that, repositioning the workforce.
On a societal level, the opportunities are in solving problems facing us today like addressing poverty or improving food security, health, safety, etc. In fact, the opportunity is so large that each person can pick their own problem to solve, whether it is big or small. Each one, pick one. I invite the readers to take this challenge! Our imagination and the will to bring about positive changes are the only real limits with these rapidly evolving technologies by our side.
What kind of challenges or opportunities are you looking to tackle next?
I’m driven by the exact same opportunity I described earlier—to solve complex, large-scale problems that have a meaningful impact. I’d be excited to lead an organization’s technological transformation, leveraging AI and other emerging technologies to redefine what’s possible. This means building and mentoring world-class engineering teams and shaping a forward-looking technology strategy that aligns with and drives business objectives. I want to take on challenges that require both strategic foresight and the ability to execute with precision, ultimately creating sustainable value and competitive advantage for a large, dynamic company.
Finally, what is the one skill you think is necessary for leadership?
Integrity. It is non-negotiable.
Bio: Deven Panchal
Deven Panchal is an accomplished leader and widely recognized expert in the fields of AI, large-scale software systems, networking, and open source. He is currently a principal software engineer with AT&T Labs, United States.
Panchal helped to build the world’s first open source AI marketplace, was involved with AT&T’s 1st 5G VRAN/CRAN trials, and has built solutions that power many of today’s global telco networks. His work has directly benefited millions of people around the world and has been featured in international media including Forbes, TechCrunch, Light Reading, and TelecomTV. He is the author of many international publications and open source projects, and has filed U.S. patent applications.
He has been honored for pioneering new directions in the fields of AI and networks, including awards such as AI Innovator of the Year by CNBC TV18; Best Paper and Best Presenter awards; the AI100 Generative AI Leader award; and D CEO’s Emerging Leader Award.
Panchal is also active in many international conferences and journals, including as a keynote speaker, a reviewer, and a member of TPCs and editorial boards. He participates in many IEEE standards committees in the areas of AI, generative AI, and safe and trustworthy AI. His research contributions have led to groundbreaking advancements in AI democratization and network automation, which led to his being elected to a senior member of IEEE and a full member of Sigma Xi.
Panchal is also VP of Intellectual Property and Patents at AT&T Innovation Network, where he helps drive innovation and emerging tech education programs for the approximately 200,000 employees in the company. He graduated with a master’s in electrical and computer engineering from Georgia Tech and has completed certifications from the Wharton School and Stanford University.
Dig Deeper
To learn more about Panchal and his work,
Each week over the next few months, Tech News will highlight different Top 30 honorees. For a full list, see Computing’s Top 30 Early Career Professionals for 2024.
In addition to Computing’s Top 30, IEEE Computer Society offers many other awards; to read about the honors and the honorees—and perhaps nominate an impactful professional in your life—visit the IEEE CS Awards page.
