AI has become an increasingly common tool for identifying novel bioactives and uncovering new ingredient opportunities. For nutrition industry stakeholders, however, R&D is only one stage of a much longer process that includes substantiating claims, navigating formulation challenges and determining whether a concept is commercially viable.
Last month, NutraIngredients reported on California-based biotech company Brightseed’s expansion of its proprietary bioactive dataset to 21 million compounds, highlighting the growing role of data and AI in ingredient discovery. The company’s latest platform, Hummingbird, shifts the focus beyond discovery and toward development-stage decision-making.
NutraIngredients spoke with Lee Chae, PhD, co-founder and CEO of Brightseed, about how he sees AI evolving beyond ingredient identification and what that could mean for ingredient manufacturers, suppliers and product developers.
NIU: Brightseed describes Hummingbird as an “agentic AI” system. What does that actually mean in practice for a team developing ingredients or finished products?
Lee Chae, PhD: Agentic AI means Hummingbird is designed to actively support decision-making and execution across development workflows—not simply generate information or respond to prompts like a general-purpose AI assistant.
For a life science or product development team, that means Hummingbird can help evaluate biological relevance, surface supporting evidence, identify potential risks or gaps, prioritize higher-confidence concepts, and organize development pathways into structured, actionable outputs.
The goal is not to replace scientific expertise or make decisions autonomously. The goal is to help teams arrive at higher-confidence, “decision-grade” recommendations earlier in the innovation process so they can focus resources on the opportunities with the greatest probability of success.
NIU: How does Hummingbird build on your Forager discovery engine, and what new capabilities does it introduce at the development stage?
Lee Chae, PhD: Forager and Hummingbird serve distinct but connected roles within Brightseed’s innovation platform.
Forager is our AI-powered discovery engine. It analyzes Brightseed’s proprietary biological dataset—which now spans more than 21 million bioactives—to identify novel compounds, biological mechanisms, and potential health opportunities at a scale not possible through traditional research methods.
Hummingbird builds on those discovery insights and applies agentic intelligence at the development stage. It helps teams evaluate feasibility, assess risk earlier, prioritize concepts, organize evidence, and advance promising opportunities toward commercialization.
In simple terms, Forager helps identify what’s biologically interesting and potentially valuable, [while] Hummingbird helps determine what is most actionable, defensible, and commercially viable.
NIU: What differentiates Hummingbird from general-purpose AI tools, particularly in how it uses proprietary, experimentally validated biological data?
Lee Chae, PhD: The biggest difference is that Hummingbird was built specifically for life sciences innovation and operates within Brightseed’s proprietary scientific architecture.
Most general-purpose AI tools are trained primarily on public internet or literature data. While they can be useful for broad information retrieval, they often lack biological depth, traceability, scientific context, and experimentally validated datasets needed for high-confidence scientific decision-making.
Hummingbird is grounded in Brightseed’s proprietary bioactive dataset and years of biological modeling, validation work, and commercial experience. That includes experimentally validated relationships between compounds, mechanisms, and health outcomes.
NIU: How does the platform shift decision-making earlier in the R&D process, and what types of late-stage risks or failures can it help companies avoid?
Lee Chae, PhD: One of the biggest challenges in health innovation is that critical risks often surface too late—after significant time, capital, and organizational effort have already been invested.
By connecting discovery insights with development intelligence earlier in the process, the platform helps teams evaluate opportunities sooner and with greater context.
That can help identify issues related to:
- Weak biological rationale
- Insufficient evidence depth
- Claims substantiation gaps
- Low differentiation potential
- Formulation complexity, and
- Commercial feasibility concerns
NIU: For ingredient manufacturers and suppliers, where will this have the most immediate impact—R&D efficiency, cost reduction, time-to-market, or product differentiation?
Lee Chae, PhD: We think the immediate impact will be in improving decision quality and accelerating innovation confidence.
That can certainly translate into greater R&D efficiency and faster development cycles, but the larger opportunity is helping companies identify and advance more differentiated, scientifically defensible products.
In today’s market, companies are under increasing pressure to substantiate claims, stand out in crowded categories, and make smarter innovation investments. Platforms like Brightseed help shift innovation from a fragmented, trial-and-error process into a more connected, evidence-driven capability.
Ultimately, we believe the biggest value is increasing the probability that innovation efforts lead to commercially viable outcomes.
NIU: Anything else to add?
Lee Chae, PhD: What’s exciting about Hummingbird is that it represents a broader shift happening across health and life sciences.
For years, companies have used disconnected tools and workflows to manage increasingly complex innovation challenges. We believe the future will belong to AI-native platforms that connect discovery, development, and decision-making into a continuous model.
Hummingbird is an important step in that direction—not as a replacement for scientists or product developers, but as a system designed to help them move faster, reduce uncertainty, and make better decisions earlier.
That’s ultimately what we believe will transform how health innovation gets done.
