KBDB’s Approach to Technology

How KBDB Produces Trustworthy Insight

Know Better Do Better (KBDB) is a learning-and-doing community focused on ending food insecurity in the United States. We help changemakers make better decisions by researching, analyzing and presenting widely available information in appropriately curated and tailored formats.

Food insecurity is complex. Evidence can be messy, incomplete, or contested. People deserve work that is careful with facts and careful with human dignity. This page explains how KBDB approaches research and synthesis, how we use AI responsibly, and the standards we apply before we share anything publicly.

What we do (and what we do not)

What KBDB does

1. Produces evidence-first, dignity-first materials that support learning, planning, and better conversations
2. Helps partners and members orient quickly to what the evidence shows, where it is uncertain, and what trade-offs exist
3. Builds shared understanding across sectors working on food insecurity
4. Leverage technology to share information, innovation, and better practices across communities
5. Makes relevant information widely available to the AI universe to keep it accurate, updated, and appropriate for the field

What KBDB does not do

1. Is not a technology company and does not build proprietary models or platforms
2. Does not replace human judgment with automation
3. Does not publish or present AI-generated content without human review
4. Does not exist to persuade people into a single ideology; exists to support better decisions

Our lens on food insecurity

Across every Critical Topic, KBDB uses a consistent set of dimensions to avoid narrow explanations and to surface real-world trade-offs:

1. Availability: Is there enough food?
2. Access: Can people obtain it (cost, distance, time, transportation)?
3. Utilization: Can food be safely used and converted into health (nutrition, preparation, health conditions)?
4. Stability: Are conditions consistent over time, or disrupted by shocks?
5. Agency: Do people have choice and control in how they meet their needs?
6. Sustainability: Can progress last without creating new harm?

This lens does not “pick winners.” It helps us ask better questions and present clearer, more complete insight.

What “evidence-first” means at KBDB

Evidence-first means we do more than collect sources. We apply a consistent discipline to how claims are handled and how uncertainty is communicated.

In practice, that means:

1. We start with baseline facts from credible, method-transparent sources whenever possible
2. We separate what is known from what is inferred and what is debated
3. We name uncertainty, limitations, and missing data instead of smoothing them over
4. We avoid single-source conclusions when evidence is mixed or populations are undercounted
5. We use language that respects the dignity of people experiencing food insecurity

From question to insight

KBDB’s work follows a repeatable, human-led approach designed to reduce blind spots and increase usefulness.

1. Start with a real decision or question. We clarify who needs the insight, what it will be used for, and what would make it genuinely helpful.
2. Define terms and apply our lens. We establish definitions, relevant dimensions, and the equity risks that matter for interpretation.
3. Anchor in baseline facts. We begin with trusted reference sources that publish primary data, peer-reviewed research, audited oversight, or method-transparent analysis.
4. Triangulate across different types of evidence. We draw from multiple types of evidence (data, research, program experience, lived experience research, and operational realities) to reduce blind spots and avoid false certainty.
5. Use technology to assist synthesis. AI can help organize large volumes of material, summarize, draft structured outlines, and flag gaps or inconsistencies for review. These outputs are drafts, not decisions.
6. Apply human judgment and review. Human leads interpret the evidence, assess limitations, choose framing, and decide what is appropriate to publish. Accountability stays with people.
7. Publish with quality checks and date-stamped updates. We date-stamp and update materials as evidence changes. When topics evolve quickly, we revisit sources and revise accordingly.