SEO / Portfolio / Public Site
General-Purpose Deep Research Brief
Report summary
Because no specific research topic was provided, this report treats the topic as unspecified and reframes the task as a portfolio-selection brief : a way to choose the best research domain before committing to a literature review, pilot, or grant application. On that basis, the strongest general-pur
Key topics
- SEO / Portfolio / Public Site
- SEO
- Portfolio
- Public Site
- AI
- Agentic Web
- Privacy
- Research Archive
- Strategy
Research provenance
For citation, use the report title and canonical URL. Archival presence does not establish authorship or promote report statements into portfolio evidence.
Source availability: 60 citation markers in the source export have no recoverable source links. Those markers are omitted from this reader; any supplied bibliography and ordinary links remain. Check the original sources before relying on the cited claims.
This page renders the archived Markdown as safe, formatted HTML. It is background research and does not become a portfolio claim without evidence review.
Full report
On this page
Executive summary
Because no specific research topic was provided, this report treats the topic as unspecified and reframes the task as a portfolio-selection brief: a way to choose the best research domain before committing to a literature review, pilot, or grant application. On that basis, the strongest general-purpose starting points are usually AI safety and frontier model evaluation, cybersecurity and software supply-chain defense, healthcare delivery and digital health, biotech and programmable medicine, and renewable energy and storage. Those domains combine large public datasets, active regulation or standards activity, and a dense recent research pipeline, including the NIST AI RMF, the EU AI Act, the growth of benchmark ecosystems such as HELM and SWE-bench, formalized software security guidance like NIST SSDF, TEFCA-enabled health-data exchange, the first FDA-approved CRISPR-based therapies, and rapidly expanding renewables and EV data ecosystems.
If the goal is near-term, evidence-rich, lower-friction research, the best “first pilot” candidates are typically cybersecurity, AI evaluation, healthcare digital operations, education technology, and urban mobility, because they offer faster experimental cycles and lower capital intensity than quantum, large-scale climate hardware, or space infrastructure. If the goal is high novelty with longer payback, quantum computing, space commercialization, and some forms of industrial climate tech remain attractive, but they generally require more specialized expertise, longer validation horizons, and larger budgets.
A practical default is to run a 12-month pilot with a narrow, measurable question; use official datasets and original papers first; and force periodic “continue / pivot / stop” decisions based on impact, tractability, data access, and policy relevance. This report provides the selection matrix, domain-specific briefs, a prioritized research-plan template, a source and search strategy, plus ready-to-use templates for a one-page literature review and a two-page grant proposal.
Scope and assumptions
This brief assumes the following are unspecified: target sector, geography, institution type, success metric, preferred methods, budget ceiling, and whether the desired output is academic, commercial, policy, or philanthropic. Therefore, the document emphasizes domains that are broad enough to matter and concrete enough to pilot.
A separate uploaded note in this conversation appears to describe a different, API-reliability research task and was treated as out of scope for this multi-domain brief. The file remains available here: Pasted markdown.
The domain ratings below are heuristic portfolio scores, not econometric forecasts. “Impact” estimates societal or strategic importance; “feasibility” estimates how quickly a competent team can produce useful evidence; “cost” is the expected pilot-study burden; “timeline” is time to first credible signal; and “novelty” estimates how differentiated a new project could still be in 2026.
Domain comparison
Suggested domains and rationale
| Domain | Why it belongs on the shortlist | Typical decision trigger |
|---|---|---|
| AI safety and frontier model evaluation | Regulation, standards, and benchmarking are moving quickly; the research frontier still has major gaps in evaluation, control, transparency, and real-world misuse measurement. | Choose if you want fast-moving, policy-relevant, benchmark-heavy work. |
| Climate tech and carbon management | Climate-risk evidence is mature, but deployment, monitoring, and industrial decarbonization remain under-built; the opportunity is large and cross-sectoral. | Choose if you want infrastructure, emissions, MRV, or industrial transition work. |
| Biotech and programmable medicine | Gene editing, AI-driven biology, and large-scale open biological data are creating unusually strong research leverage. | Choose if you want wet-lab adjacency with strong computational leverage. |
| Renewable energy and long-duration storage | Deployment data are rich, market demand is global, and storage/grid integration remain major bottlenecks. | Choose if you want techno-economic modeling with policy relevance. |
| Quantum computing and quantum networking | The field remains high-novelty; recent milestones in logical qubits, utility-scale systems, and post-quantum readiness make it strategically important despite slower payoff. | Choose if you want frontier technical depth and can tolerate slower pilots. |
| Cybersecurity and software supply-chain defense | Standards, vulnerability data, and exploit-prioritization infrastructure are mature; pilots can produce measurable operational value quickly. | Choose if you want strong public data and rapid evidence-to-practice cycles. |
| Healthcare delivery and digital health | Interoperability, digital public health, and large clinical data resources make it possible to study operational improvement rather than only models. | Choose if you want measurable human outcomes with real implementation constraints. |
| Education technology and learning recovery | Learning-loss evidence is strong, and GenAI raises urgent design and governance questions in classrooms and research. | Choose if you want relatively low-cost pilots with visible public impact. |
| Supply chain resilience and trade infrastructure | Resilience questions remain policy-critical; trade, tariff, and GVC datasets make cross-country analysis tractable. | Choose if you want macro-to-operational analysis with strong policy framing. |
| Space commercialization and Earth observation | Commercial LEO, lunar economy work, and open Earth-observation archives create a dual-use research space spanning economics, logistics, and sensing. | Choose if you want strategic technology plus open geospatial data. |
| Urban mobility and transport electrification | EV adoption, charging systems, and multimodal travel data create a strong applied research field with high city-scale relevance. | Choose if you want deployable pilots with transportation and climate co-benefits. |
Comparison matrix
| Domain | Impact | Feasibility | Pilot cost | Time to first signal | Novelty |
|---|---|---|---|---|---|
| AI safety and frontier model evaluation | Very high | High | Medium | Medium | High |
| Climate tech and carbon management | Very high | Medium | Medium to high | Medium | High |
| Biotech and programmable medicine | Very high | Medium | Medium to high | Medium | Very high |
| Renewable energy and long-duration storage | Very high | High | Medium | Medium | Medium |
| Quantum computing and quantum networking | High | Low to medium | High | Slow | Very high |
| Cybersecurity and software supply-chain defense | Very high | Very high | Low to medium | Fast | Medium |
| Healthcare delivery and digital health | Very high | High | Medium | Medium | Medium |
| Education technology and learning recovery | High | Very high | Low to medium | Fast | Medium |
| Supply chain resilience and trade infrastructure | High | High | Low to medium | Fast to medium | Medium |
| Space commercialization and Earth observation | High | Medium | Medium to high | Medium | Very high |
| Urban mobility and transport electrification | High | High | Medium | Fast to medium | Medium |
xychart-beta
title "Illustrative weighted priority score"
x-axis ["Cyber","AI eval","Healthcare","Biotech","Renewables","Climate","Education","Urban","Supply chain","Space","Quantum"]
y-axis "Score" 0 --> 100
bar [86,84,82,80,79,76,74,72,70,66,62]Default prioritization if no other preferences are specified: cybersecurity, AI evaluation, healthcare delivery, biotech, and renewables/storage. That ordering reflects the best balance of public data access, short pilot cycles, and policy or operational relevance.
Domain briefs
Questions, developments, organizations, and data
| Domain | Key questions | Recent major developments in roughly the last five years | Representative organizations and research groups | Primary data sources and datasets |
|---|---|---|---|---|
| AI safety and frontier model evaluation | Which evaluations predict real-world failure or misuse? How should agentic capability, persuasion, robustness, and post-deployment drift be measured? What transparency is sufficient for downstream safety and governance? | NIST released the AI RMF; the EU AI Act entered into force in July 2024; government-backed frontier testing expanded through AISI; benchmark ecosystems became more standardized and broader, including HELM and SWE-bench. | NIST; UK AISI; Stanford CRFM/HAI; benchmark maintainers such as HELM and SWE-bench. HELM explicitly names Percy Liang and Rishi Bommasani among its authors. | HELM benchmark outputs and prompts; SWE-bench leaderboards and task sets; public transparency indices; public model documentation and evaluation reports. |
| Climate tech and carbon management | Which interventions reduce emissions fastest per dollar? How should carbon removal, methane reduction, and industrial decarbonization be monitored and verified? Which technologies are cost-effective in hard-to-abate sectors? | The IPCC AR6 Synthesis Report sharpened the evidence base; the Global Carbon Budget moved to a dedicated release site; renewable-capacity tracking improved; industrial decarbonization, carbon-management, and MRV remain active commercialization fronts. | IPCC; IEA; IRENA; Global Carbon Project; national labs and climate-observation programs. | Global Carbon Budget; IRENA capacity statistics; NASA Earthdata; IEA report datasets and explorers. |
| Biotech and programmable medicine | Which therapeutic modalities are most de-risked for near-term trials? How should AI structure prediction, genomic variant interpretation, and biobank data be combined? Where are the translational bottlenecks? | FDA approved the first gene therapies for sickle cell disease, including the first CRISPR-based therapy; AlphaFold DB scaled to more than 200 million predicted structures; clinical-variant and biobank resources continued to expand. | FDA; NIH/NCBI; EMBL-EBI/DeepMind; UK Biobank; clinical genomics communities around ClinVar and GenBank. | AlphaFold DB; ClinVar; GenBank; UK Biobank; trial registries and FDA reviews. |
| Renewable energy and long-duration storage | Where are the highest-value storage and flexibility gaps? Which grid, pricing, or siting barriers matter most? Which chemistries or system designs have the best regional fit? | IEA’s renewables outlook extended to 2030; IRENA published decade-long renewable-capacity statistics through 2024; EV and charging datasets expanded; infrastructure-planning tools became more accessible. | IEA; IRENA; NREL; grid operators; public charging-infrastructure analysis groups. | IRENA Renewable Capacity Statistics; IEA Global EV Data Explorer and outlook datasets; EV-infrastructure planning tools such as EVI-Pro. |
| Quantum computing and quantum networking | Which applications are credible before full fault tolerance? What error-correction milestones matter most? How should post-quantum migration and quantum advantage be tracked separately? | NIST continues to frame quantum science and post-quantum cryptography as national priorities; Google reported a logical-qubit milestone in Nature; IBM declared an “era of quantum utility” and expanded hardware and software roadmaps. | NIST; IBM Quantum; Google Quantum AI; university joint institutes such as JQI, JILA, and QuICS. | NIST quantum resources; vendor hardware data and roadmaps; benchmark and calibration studies; open papers in quant-ph. |
| Cybersecurity and software supply-chain defense | How should teams prioritize patching under limited capacity? Which controls measurably reduce exploitable risk? How should SBOMs, build provenance, and exploit prediction work together? | NIST finalized SSDF v1.1 in 2022; NVD continues to standardize vulnerability data and CVSS support; EPSS now provides daily exploit-probability estimates for CVEs. | NIST; NVD; FIRST EPSS community; software security practitioners and supply-chain security researchers. EPSS publicly identifies Jay Jacobs and Stephen Shaffer among SIG chairs and contributors. | NVD feeds and APIs; EPSS CSV/API; internal or public SBOM/provenance artifacts; KEV-like exploited-vulnerability lists where accessible. |
| Healthcare delivery and digital health | Which digital interventions improve outcomes or reduce administrative burden? How should nationwide interoperability change workflow design? What is the right balance between prediction and operations research? | TEFCA continues to structure nationwide exchange; WHO expanded global digital-health coordination, including AI-related work and digital public infrastructure; MIMIC-IV and HCUP remain key structured clinical data assets. | ONC; WHO; AHRQ; PhysioNet; health-system analytics groups. MIMIC-IV names Alistair Johnson and Leo Anthony Celi among dataset authors. | ONC datasets; MIMIC-IV; HCUP; CMS open data; public-health digital standards resources. |
| Education technology and learning recovery | Which interventions improve learning recovery with limited teacher burden? How should GenAI be used without worsening inequity or distraction? What should the outcome metric be: test scores, persistence, engagement, or metacognition? | PISA 2022 showed record OECD drops in mathematics and large reading declines from 2018 to 2022; NAEP long-term trends showed severe weakness in reading habits and achievement; UNESCO published global guidance for GenAI in education and research. | OECD/PISA; NAEP/NCES; UNESCO; learning-science and curriculum-design communities. UNESCO’s guidance is authored by Fengchun Miao and Wayne Holmes. | PISA microdata and reports; NAEP releases; UNESCO education data and guidance. |
| Supply chain resilience and trade infrastructure | Which dependencies are truly critical? When does diversification outperform reshoring? Which node, corridor, or product-level concentrations create the largest systemic risk? | The New York Fed’s GSCPI has become a standard reference for pressure tracking; WITS now integrates trade, tariffs, NTMs, and GVC data; resilience debates increasingly focus on diversification rather than simple reshoring. | New York Fed; World Bank WITS; WTO/UNCTAD-linked trade-data providers; OECD trade-policy communities. | GSCPI; WITS trade, tariff, NTM, and GVC data; UN/World Bank trade sources linked through WITS. |
| Space commercialization and Earth observation | Which business models survive beyond launch? Where do public-private partnerships create the most value? How can Earth-observation data create commercial or policy value without requiring a new satellite? | NASA is explicitly transitioning toward a commercial LEO model and broader public-private partnerships; FAA AST continues to regulate and promote commercial launch/reentry activity; NASA Earthdata continues to expand open Earth-observation access. | NASA commercial-space programs; FAA AST; NASA Earthdata/DAAC network; Earth-observation analytics groups. | NASA Earthdata; DAAC archives; launch and licensing records; mission and payload program data. |
| Urban mobility and transport electrification | How much charging, storage, curb, and modal integration will cities need? Which interventions reduce travel time, emissions, and inequity together? How should multimodal demand be forecast under EV adoption? | EV outlooks now cover affordability, batteries, charging, and projections to 2030; FHWA’s NextGen NHTS is moving toward more frequent travel-behavior products; EVI-Pro-like tools support charging buildout analysis. | IEA; FHWA; national transportation labs and charging-infrastructure modelers. | IEA EV data; NHTS and NextGen OD products; charging-infrastructure projection tools. |
Gaps, pilot ideas, ethics, and resource estimates
| Domain | Open problems and research gaps | Potential high-impact pilot or experiment | Ethical and regulatory considerations | Estimated pilot resource needs |
|---|---|---|---|---|
| AI safety and frontier model evaluation | Weak links between benchmark scores and real-world harms; sparse post-deployment monitoring; limited causal evidence on safeguards. | Build a domain-specific eval harness for one real workflow, such as software operations, research synthesis, or triage, then compare model variants and safeguard stacks. | Safety, misuse, privacy, labor displacement, transparency obligations, sector rules under the EU AI Act and related regimes. | 4–6 months, $250k–$1.2M, 4–6 FTE spanning ML eval, product instrumentation, domain expert, and governance lead. |
| Climate tech and carbon management | MRV quality, durability and permanence for removals, cost uncertainty in hard-to-abate sectors, local siting and justice trade-offs. | Build a regional industrial-decarbonization atlas combining emissions, infrastructure, energy price, and storage or removal pathways. | Environmental justice, lifecycle accounting, land use, permitting, verification standards, public acceptance. | 6–9 months, $500k–$2M for analytics-only; more if physical testing is included; 5–8 FTE with geospatial, techno-economic, and policy expertise. |
| Biotech and programmable medicine | Translation from computational promise to validated therapy; generalization across cohorts; variant interpretation bottlenecks; data-access friction. | Train and validate a variant-prioritization or protein-function scoring pipeline against one disease area using ClinVar, GenBank, and biobank-linked phenotypes. | Human-subjects protections, genomic privacy, FDA pathway strategy, dataset permissions, bias in ancestry representation. | 6–12 months, $750k–$3M depending on wet-lab work; 5–10 FTE across bioinformatics, statistics, translational science, and ethics/compliance. |
| Renewable energy and long-duration storage | Long-duration cost curves, grid integration, siting friction, interconnection queues, region-specific demand uncertainty. | Run a city or utility charging-plus-storage optimization study using public mobility and power-demand assumptions. | Grid reliability, rate design, community siting, environmental review, mineral sourcing and recycling. | 4–8 months, $300k–$1.5M, 4–7 FTE in energy systems, optimization, geospatial analytics, and stakeholder engagement. |
| Quantum computing and quantum networking | Application-market mismatch, fragile benchmarks, unclear near-term economic use cases, slow path to large-scale fault tolerance. | Evaluate one candidate use case, such as chemistry simulation or optimization, against classical baselines and hardware constraints. | Export controls, cryptographic transition risk, reproducibility, vendor lock-in, overclaiming utility. | 6–12 months, $500k–$2M, 3–6 FTE with quantum algorithms, HPC/classical baseline, and domain science expertise. |
| Cybersecurity and software supply-chain defense | Incomplete exploit data; limited evidence on which controls reduce real exploitability; SBOM quality and provenance gaps. | Build a patch-prioritization engine that combines NVD, EPSS, and asset criticality, then test whether it outperforms CVSS-only triage. | Disclosure timing, liability, logging/privacy, secure handling of asset inventories, regulatory obligations in critical sectors. | 3–6 months, $150k–$750k, 3–5 FTE in AppSec, data engineering, and operations research. |
| Healthcare delivery and digital health | Evidence often focuses on prediction rather than implementation; interoperability still does not guarantee workflow transformation; evaluation designs can be weak. | Pilot one operational use case, such as referral closure, discharge follow-up, or ED throughput, using interoperable clinical/event data and pre-registered outcome metrics. | HIPAA/privacy, clinical safety, IRB review, information blocking, interoperability standards, provider burden. | 4–9 months, $300k–$1.5M, 4–8 FTE including clinical ops, health informatics, analytics, and compliance. |
| Education technology and learning recovery | Weak causal evidence for many EdTech tools, heterogeneous classroom adoption, risk of distraction and inequity from GenAI. | Run a randomized or quasi-experimental tutoring or feedback intervention with teacher-in-the-loop controls and narrow learning objectives. | Student privacy, age-appropriate AI use, teacher autonomy, accessibility, procurement and district approvals. | 3–9 months, $100k–$800k, 3–6 FTE with learning science, product measurement, and school-partnership capacity. |
| Supply chain resilience and trade infrastructure | Criticality mapping is often too coarse; product-level substitutions are poorly modeled; resilience metrics are inconsistent. | Build a product-to-corridor dependency map for one industry and simulate diversification, inventory, and reshoring scenarios. | Confidentiality of supplier relationships, geopolitical sensitivity, trade-law implications, antitrust and procurement constraints. | 4–8 months, $200k–$1M, 3–6 FTE in trade data, network analysis, operations, and policy. |
| Space commercialization and Earth observation | Launch is crowded but downstream value capture remains uneven; many business models need stronger unit economics; EO-to-decision pipelines are underdeveloped. | Use open EO data to build a narrow decision product, such as wildfire, agriculture, water, or port-congestion monitoring, without launching hardware. | Spectrum and licensing, remote-sensing policy, national-security constraints, public procurement dependency. | 4–9 months, $250k–$1.5M, 4–7 FTE with remote sensing, product analytics, and domain stakeholder expertise. |
| Urban mobility and transport electrification | Demand forecasts can break under policy change; equity effects are under-measured; charging deployment often lags travel reality. | Combine NHTS OD products with charger-demand modeling for one metro area to identify under-served corridors and curbside gaps. | Location privacy, equity across neighborhoods, utility coordination, local permitting, curb governance. | 4–8 months, $200k–$1M, 3–6 FTE in transport modeling, GIS, utility or city policy, and community engagement. |
Research execution toolkit
A good default research program starts with a portfolio down-select, then moves to problem framing, evidence review, pilot design, data access and governance, implementation, and decision review. When details are unspecified, this sequence prevents premature commitment to a domain that is exciting but poorly scoped.
Prioritized research-plan template
| Component | What to decide | Recommended default |
|---|---|---|
| Decision criteria | What determines domain choice? | Impact, tractability, data access, implementation path, novelty, regulatory clarity, and team fit. |
| Domain down-select | How many domains survive stage one? | Cut from 10–11 domains to 3, then to 1 after two weeks of structured review. |
| Main question | What is the single highest-value answer? | One sentence, testable, bounded to one user, geography, process, or disease/asset/classroom/industry. |
| Milestones | What must be true at each gate? | Problem statement approved; literature map complete; dataset access secured; baseline established; pilot launched; interim review; final recommendation. |
| Methods | How will evidence be produced? | Mixed methods: structured literature review, benchmark analysis, stakeholder interviews, prototype or quasi-experiment, and simple decision economics. |
| Deliverables | What will be handed over? | Research memo, annotated bibliography, dataset register, reproducible analysis notebook, pilot results deck, and next-step recommendation. |
| Risk mitigation | What can fail first? | Pre-register assumptions; maintain fallback datasets; define kill criteria early; separate “research unknowns” from “access unknowns.” |
Recommended methods by stage
| Stage | Best-fit methods | Primary outputs |
|---|---|---|
| Scoping | Problem framing, stakeholder interviews, concept map | 2-page scoping memo |
| Evidence review | Systematic or semi-systematic review, benchmark scan, policy scan | Literature matrix and gap map |
| Baseline | Descriptive stats, benchmark reproduction, process mapping | Baseline report |
| Intervention design | Prototype, simulation, lab test, or quasi-experiment design | Pilot protocol |
| Pilot | A/B test, before-after, RCT, synthetic control, or case-comparison | Interim results and operational notes |
| Decision | Cost-effectiveness, robustness checks, risk review | Go / pivot / stop memo |
Twelve-month pilot timeline
gantt
title 12-month pilot plan
dateFormat YYYY-MM-DD
axisFormat %b %Y
section Select
Domain screening and scoring :a1, 2026-08-01, 30d
Final topic selection and question :a2, after a1, 15d
section Review
Literature review and source mapping :b1, after a2, 45d
Data access and governance setup :b2, after a2, 45d
section Build
Baseline analysis :c1, after b1, 45d
Pilot design and preregistration :c2, after b1, 30d
Prototype or intervention build :c3, after c2, 60d
section Test
Pilot execution :d1, after c3, 75d
Interim review and pivot gate :milestone, d2, after d1, 1d
section Synthesize
Final analysis and robustness checks :e1, after d1, 45d
Final memo, deck, and grant package :e2, after e1, 30dSearch strategy and prioritized sources
The search order should usually be: official policy or standards documents → original papers and dataset papers → public datasets and APIs → high-quality synthesis reports → only then news or vendor commentary. This is especially important in domains with fast-moving regulation, benchmark churn, or strong vendor incentives.
| Priority | Source class | Why it comes first | Examples |
|---|---|---|---|
| Highest | Official and statutory sources | They define the binding rules, program scope, and institutional vocabulary. | EUR-Lex for the AI Act, NIST frameworks, WHO strategy pages, ONC TEFCA, FAA AST, NASA Earthdata, AHRQ HCUP. |
| High | Original papers and dataset papers | They contain methods, assumptions, and often the real limitations. | HELM, SWE-bench, MIMIC-IV, Google logical-qubit paper coverage, benchmark papers, clinical dataset papers. |
| High | Primary public datasets and explorers | They enable replication and pilot design. | NVD, EPSS, WITS, NHTS, IRENA capacity data, IEA EV data, ClinVar, GenBank, UK Biobank. |
| Medium | Flagship intergovernmental reports | They provide structured synthesis and scenario framing. | IPCC, IEA, OECD, UNESCO. |
| Lower | Vendor blogs, think-tank commentaries, news | Useful for context and emerging signals, but not as first-order evidence. | Vendor roadmaps, financial press, industry commentary. |
Suggested keyword patterns
| Domain | Suggested query patterns |
|---|---|
| AI safety | "frontier model evaluation" benchmark governance, "site:nist.gov" AI Risk Management Framework, "site:eur-lex.europa.eu" 2024/1689 artificial intelligence, "HELM" standardized evaluation, "SWE-bench" verified software agents |
| Climate tech | "IPCC AR6 synthesis" mitigation pathways, "Global Carbon Budget" sector emissions, "industrial decarbonization" MRV carbon management, "renewable capacity statistics" IRENA, "Earth observation" emissions monitoring |
| Biotech | "site:fda.gov" CRISPR gene therapy approval, "AlphaFold DB" protein structure database, "site:ncbi.nlm.nih.gov" ClinVar variant interpretation, "site:ncbi.nlm.nih.gov" GenBank, "UK Biobank" disease prediction |
| Renewables and storage | "long duration storage" technoeconomic analysis, "site:iea.org" renewables outlook 2030, "IRENA renewable capacity statistics", "EV charging infrastructure" EVI-Pro, "grid flexibility" storage siting |
| Quantum | "logical qubit" surface code milestone, "site:nist.gov" quantum information science, "quantum utility" IBM roadmap, "post-quantum cryptography" standards transition, "quantum networking" standards |
| Cybersecurity | "site:csrc.nist.gov" SSDF 800-218, "site:nvd.nist.gov" vulnerability data feeds, "EPSS" exploit probability, "software supply chain provenance" SLSA SBOM, "patch prioritization" exploit in the wild |
| Healthcare digital | "TEFCA" interoperability operations, "digital health workflow improvement", "MIMIC-IV" operational prediction, "HCUP" care delivery outcomes, "WHO digital health" AI policy |
| Education tech | "PISA 2022" learning loss interventions, "NAEP long-term trend" reading math, "UNESCO generative AI education guidance", "teacher-in-the-loop tutoring", "learning recovery quasi-experiment" |
| Supply chains | "GSCPI" supply chain pressure, "WITS" trade tariff NTM GVC, "friendshoring reshoring diversification", "critical dependency mapping", "product-level supply chain resilience" |
| Space and EO | "NASA commercial space low Earth orbit economy", "FAA AST launch licensing data", "NASA Earthdata" wildfire agriculture water use, "commercial lunar payload services", "Earth observation decision support" |
| Urban mobility | "Global EV Outlook" charging affordability, "NHTS origin destination" travel behavior, "EV charging gap analysis", "multimodal accessibility electrification", "curbside charging equity" |
Templates
One-page literature review template
| Section | Target length | What to include |
|---|---|---|
| Topic statement | 2–3 sentences | The exact problem, why it matters, and what is still unspecified. |
| Research question | 1 sentence | A bounded, testable question. |
| Search strategy | 3–4 lines | Databases searched, date range, inclusion/exclusion rules, and query logic. |
| Core findings | 2 short paragraphs | What the literature agrees on; where results diverge; strongest empirical patterns. |
| Methods landscape | 1 short paragraph | Dominant methods, datasets, benchmarks, and identification strategies. |
| Gaps | 3–5 lines | What is missing, weakly measured, or still poorly causal. |
| Implications | 3–5 lines | Why the gap matters for policy, operations, commercialization, or follow-on research. |
| Proposed next study | 3–5 lines | Hypothesis, dataset, intervention or method, expected deliverable. |
| References | 5–8 items | Original papers, official reports, dataset papers only unless no primary source exists. |
Two-page grant proposal template
| Section | Target length | What to include |
|---|---|---|
| Project title and abstract | 120–180 words | Problem, intervention, domain, expected contribution, and why now. |
| Need and significance | 1–2 paragraphs | Evidence of importance; why current tools or knowledge are insufficient. |
| Specific aims | 3 short aims | Each aim should produce an observable output. |
| Background and prior work | 1 paragraph | Key literature, prior pilots, and why this project is differentiated. |
| Research design and methods | 2–3 paragraphs | Data, sample, intervention, baseline, comparison design, and analysis plan. |
| Milestones and deliverables | 1 paragraph or mini-table | What is delivered by month 3, month 6, month 9, and month 12. |
| Team and capabilities | 1 paragraph | Roles, missing expertise, partner institutions, and advisory needs. |
| Risk and mitigation | 1 paragraph | Data-access risk, ethics or regulatory risk, technical risk, adoption risk. |
| Budget summary | 4–6 lines | Personnel, data/licensing, compute/lab, participant costs, travel, overhead if applicable. |
| Impact and scaling path | 1 paragraph | How successful results would be adopted, published, or expanded. |
Minimal fill-in prompt for whichever domain you choose next
| Field | Fill-in text |
|---|---|
| Chosen domain | [domain] |
| Geography | [country / region / city / global / unspecified] |
| Main user or beneficiary | [who benefits] |
| Pilot question | [one bounded question] |
| Primary outcome metric | [one metric] |
| Secondary metric | [one metric] |
| Dataset or source of truth | [official dataset / benchmark / registry] |
| Pilot method | [benchmark / simulation / quasi-experiment / RCT / prototype] |
| Kill criterion | [what would make you stop] |
| End deliverable | [memo / paper / grant / prototype / dashboard] |
The fastest path from this brief to a real project is to pick one of the Tier 1 domains, narrow it to one operational question, and then run the 12‑month pilot plan above using primary sources first. The domains most likely to produce a rigorous, fundable study with unspecified constraints are cybersecurity, AI evaluation, healthcare digital operations, biotech, and renewables/storage.