Discover the affordable innovators powering the next wave of artificial intelligence.
AI Has a Momentum Problem & That’s an Opportunity
Artificial intelligence isn’t a passing trend; it represents a foundational shift in how technology is built and applied. From logistics and manufacturing to language processing and national security, AI is steadily reshaping how systems operate and decisions are made.
Yet most investor attention remains concentrated on a small group of mega-cap names. Beneath the surface, a broader ecosystem of companies is working on the tools, infrastructure, and applications that make AI possible — often with far less visibility.
That gap may be where opportunity exists.
While some AI stocks have already experienced dramatic runs, others remain relatively affordable. Many trade below $20 and are still in earlier stages of adoption, with developing partnerships, pilot programs, or niche use cases tied to long-term AI trends.
This report highlights five public companies under $20 that may be worth watching. They are not headline-driven hype plays. Instead, they operate in areas such as power management, voice automation, defense analytics, edge computing, and data-center infrastructure — parts of the AI stack that tend to matter more as adoption scales.
Whether you’re looking to broaden your AI exposure or simply understand where enabling technologies may be heading next, these names offer a practical place to begin deeper research.
1. Navitas Semiconductor (NASDAQ: NVTS)
AI Power Management at the Core of Data Growth
Navitas Semiconductor focuses on power technologies based on gallium nitride (GaN) and silicon carbide (SiC) — materials often associated with higher efficiency and better thermal performance than traditional silicon.
As AI workloads expand, especially in data centers, power delivery and energy efficiency are becoming increasingly important considerations. Navitas has positioned its technology toward these higher-density, higher-power environments, where efficiency gains can compound at scale.
Why it may be worth researching:
-
Power efficiency is becoming a meaningful constraint in AI infrastructure, not just compute capacity.
-
Navitas’ focus on next-generation power architectures aligns with broader industry discussions around data-center scalability.
-
Its technology sits upstream of AI compute, potentially benefiting from infrastructure build-outs regardless of which models dominate.
What to watch:
-
Adoption of new power delivery standards tied to AI data-center design.
-
Evidence that advanced power architectures move from concept to broader deployment.
-
Any signals that hyperscalers or ecosystem partners increase commitment to higher-efficiency power systems.
Bonus insight:
As AI systems scale, power delivery may quietly become as critical as processing itself — making efficiency-focused suppliers worth monitoring.
2. SoundHound AI (NASDAQ: SOUN)
Voice AI for the Real World
SoundHound develops voice and conversational AI systems designed for practical, real-world environments — including restaurants, vehicles, and customer-service interactions.
Unlike general-purpose AI models, SoundHound’s platform is oriented toward embedded, task-specific voice applications, where speed, accuracy, and reliability matter more than novelty. That focus may position it differently as businesses explore automation that directly affects customer experience.
Why it may be worth researching:
-
Voice interfaces remain one of the most natural human-computer interaction layers.
-
SoundHound’s emphasis on enterprise and embedded use cases may reduce dependence on consumer AI hype cycles.
-
Long-term agreements and repeat deployments could signal durability beyond pilot programs.
What to watch:
-
Expansion of voice AI deployments in restaurants, automotive platforms, or other high-volume environments.
-
Evidence of improving monetization per deployment rather than just headline customer counts.
-
Broader enterprise acceptance of voice as a core interface, not a novelty feature.
Bonus insight:
If voice becomes a standard layer across physical and digital experiences, companies already operating in live environments may have an advantage.
3. BigBear.ai (NYSE: BBAI)
AI, Defense, and Decision Intelligence
BigBear.ai operates at the intersection of artificial intelligence, analytics, and government decision-making, with a particular focus on defense and national security applications.
Its tools are designed to help organizations analyze complex data sets, plan operations, and support decision-making in environments where uncertainty and scale are persistent challenges.
Why it may be worth researching:
-
Defense and government agencies remain steady adopters of advanced analytics and AI-driven planning tools.
-
BigBear’s positioning is tied more to long-term institutional demand than consumer sentiment.
-
The company provides exposure to AI use cases where reliability and accountability matter as much as performance.
What to watch:
-
Stability and diversification of government contract revenue.
-
Progress toward platform-based offerings versus project-specific work.
-
Signals that decision-intelligence tools gain traction beyond initial deployments.
Bonus insight:
Defense-oriented AI tends to evolve more slowly — but it can also be stickier once embedded into planning and operational workflows.
4. GSI Technology (NASDAQ: GSIT)
Chips Built for AI at the Edge
GSI Technology develops specialized processors designed for edge AI environments, where power, latency, and physical constraints limit the usefulness of traditional cloud-based compute.
Its approach centers on associative processing, a compute-in-memory architecture intended to accelerate certain AI tasks while reducing energy consumption — a potential advantage in constrained environments such as defense systems, sensors, or remote platforms.
Why it may be worth researching:
-
Edge AI is a growing category distinct from large-scale data-center AI.
-
Compute-in-memory approaches may offer efficiency advantages for specific workloads.
-
Government and research programs often act as early adopters of non-traditional architectures.
What to watch:
-
Validation of performance claims in real-world or pilot deployments.
-
Follow-on work from government or defense research programs.
-
Evidence that edge-focused AI systems move from experimentation to scaled use.
Bonus insight:
Not all AI workloads belong in the cloud. As edge use cases expand, specialized architectures may play a larger role.
5. POET Technologies (NASDAQ: POET)
Addressing One of AI’s Emerging Bottlenecks: Data Movement
POET Technologies focuses on optical interposer technology, which aims to improve how data moves between components inside high-performance systems.
As AI models grow and data-center traffic increases, data movement — not just computation — is becoming a limiting factor. POET’s platform integrates multiple photonic components into a single package, potentially reducing power use and complexity while increasing bandwidth.
Why it may be worth researching:
-
Optical interconnects are increasingly discussed as a necessity for next-generation AI systems.
-
POET’s technology targets a known scaling challenge rather than a speculative application.
-
The company operates as an enabler rather than a competitor to major chip designers.
What to watch:
-
Adoption of higher-speed optical modules in AI data centers.
-
Design wins or early commercial deployments tied to AI networking.
-
Signs that data-movement constraints become a primary focus for infrastructure investment.
Bonus insight:
As AI systems scale, moving data efficiently may matter as much as processing it — creating potential opportunities for infrastructure enablers.
Opportunity Below the Surface
AI is already influencing nearly every sector, but many investors focus on the most visible layers: models, platforms, and mega-cap hardware providers.
The companies in this report offer a different lens — smaller, more focused businesses working on specific constraints within the AI ecosystem, from power and voice interfaces to edge compute and optical connectivity.
They are not guaranteed winners. But they may offer useful insight into where the next set of challenges — and solutions — in AI development could emerge.
And for many investors, that’s where meaningful research begins.
Disclaimer
This content is for informational purposes only and should not be construed as investment advice. The companies mentioned are examples of potential research opportunities and are not endorsements or recommendations to buy or sell any security. Investing involves risk, including the possible loss of principal. Always conduct your own due diligence and consult with a licensed financial advisor before making any investment decisions.
