Following Part 1: What’s Working in Federal Contracting, this second installment focuses on how disciplined leaders are using AI to accelerate early-stage growth activities, strengthen go or no-go decisions, and build structured workflows that keep results reliable.

The questions below came directly from executives who attended our Deltek webinar on practical AI for government contractors.

 AI Prompts

Question: Can you provide tips for AI prompts to help speed up finding and reviewing government solicitations. Fed, state, and local?

Most BD teams spend hours manually scanning portals, reading attachments, and building spreadsheets. If teams can use AI to handle the research and summarization of opportunities, it allows you to focus more on decisions and strategy versus time on administrative research.

Opportunity Identification

• Prompt AI as a market analyst: “You are a public-sector market research analyst. Build a table of upcoming opportunities in [NAICS/PSC] from SAM.gov, GovWin, and relevant agency portals over the next 24 months. Columns: customer, vehicle, est. value, competition type, set-aside, due date, link, incumbent (if known), three capture risks. Include a source link or file/page for each row.”
• Add further context with prompts such as, “Using USAspending and FPDS, list the top five programs funding this work, the last three years of awards, and the spend trend. Provide links or report IDs so we can verify.”

Opportunity Review

• To triage attachments without having to read all documents, drop in the RFI, sources sought, or the past RFP and prompt: “From the RFI/sources sought/past RFP, pull the 5–10 items that would screen us in or out. past performance, clearances, key personnel, facilities, geographic limits. Quote the exact line and map to Section L/M where applicable.”

Go/No-Go Decisions

• Use existing opportunity qualification criteria into your AI and prompt, “Using our qualification criteria, score this for Validity (is it a real deal that has been funded), Viability (will this result in tangible revenue for us?), and Value (does the value help us meet our revenue objectives?) based on our vehicles, past performance, clearances, and customer access. Return a one-page read-out plus a table with five gating risks and one mitigation each. Recommend: advance, shape, or drop?”

Keep it Reliable

It will be important to continually test and refine your prompts and assess the results for accuracy. While using GenAI may speed up the process to find and review opportunities, your staff should double and triple check the results to ensure they are up to date and relevant to your growth objectives.

• Require citations to the source document
• Test your prompts on old RFPs before using them on live opportunities
• Keep a checklist: Do we have Section L/M mapped? Issues list with owners? Gaps tied to teaming or no-bid?

For state and local, swap federal portals for your state’s procurement site or BidNet as the same prompts should work.

The goal is to allocate your time and judgement on things that matter instead of spending weeks constantly searching for possible opportunity fits.

Building Your Own Library

Questions: Is there an open‑source list of proven successful “Prompt Templates” or “Engineered Prompts” for everyday capture and BD?

There isn’t a single authoritative or “GovCon-only” library yet. The most reliable approach is to build your own internal prompt library aligned with your business development process, evaluation factors, and company tone.

How to Build Your Own Library

• Start with roles: Create role prompt to uncover capture strategies or proposal approaches. For example, “a 30-year capture strategy veteran”, “a former SSEB evaluator for [Agency]”, “a Red Team trained proposal strategist who has won 90% of their deals” (ok that one I made up, but feel free to use!).
• Build in Context: Provide your challenge, situation, or goal. The more context, the better the output. For example, “Our company is aiming to double revenue in five years primarily through organic growth” or “Historically, we’ve pursued unqualified opportunities, and my task is to find, qualify, and pursue higher PWin deals.”
• Develop prompts for recurring tasks or actions: Document a list of tasks that your team frequently ask for such as agency profiles, decision-makers, capture strategy summaries, or evaluator scoring breakdowns. Make sure to store these as templates so your team doesn’t have to reinvent them every time.
• Standardize tone and output: Specify how you want the information delivered. For example
  • Tone: Formal vs. conversational; executive summary vs. analyst detail.
  • Style: Concise bullet list, proposal paragraph, or slide-ready summary.
  • Output: Define structure such as tables for compliance checks, narratives for strategy briefs, or checklists for capture actions.

Manage Your Library Like a Knowledge Base

Include templates for common workflows such as requirement shredders, resume mappers, past performance selectors, PTW scaffolds, and strengths builders. Keep this in a secure workspace with version-control. Remove or retire unused prompts into a separate section, so people can make note to not use them again. The key is to build a working toolkit that reflects your process and culture.

Public prompt libraries exist, but none are tailored for federal capture. Use them for inspiration, not as templates. For structure and risk controls, reference NIST’s Generative AI Profile, GSA’s AI guidance, and DAU’s AI resources.

Customer Intel

Question: Any key insights that will help us learn more about the customer’s main areas of focus, IT challenges, and capabilities they expect from the vendor?

AI can help you extract and synthesize what matters most to your target customer. You can gather facts from official sources, spending data, and opportunity updates, then translate those insights into strengths that sharpen your capture strategy. Always verify the results before acting, ensuring it captures your tone and actual approach.

Follow a logical sequence of prompts to keep the workflow efficient and avoid overloading one request. A multi-agent process usually delivers better, more complete results

How to Structure Your Workflow

1. Agency Executive Summary: Request a snapshot of the agency or customer before diving too deeply.

Prompt: You are an experienced federal market researcher who leaves no stone unturned. Using information I paste from agency strategic plans, acquisition forecasts, IG reports, budget justifications, and recent press releases, create a one to three page brief covering mission, leadership priorities, top programs, vehicles, spend trend, hot buttons, and pitfalls. Provide the source for each data point.

2. Award context: Once you have a picture of the agency, narrow your research to overall spend and contract vehicles to show what the customer actually buys.

Prompt: You are a former OMB executive overseeing agency budgets and now a federal research subject matter expert. From USAspending and FPDS (or Deltek GovWin), summarize the last three years of awards for [agency/mission area]. List top programs, vehicles used, prime contractors, and contract types. Show year-over-year spend and note any clear shifts. Include links or report IDs we can verify.

Note: Deltek GovWin provides an excellent summary of this information if you are a GovWin subscriber. I will still run the above prompt to cross-compare against GovWin and other data sources.

3. Opportunity capture snapshot: Assuming you have identified one or more opportunities for your pipeline, run a high-level capture analysis to give your team some context as to why this may be a viable opportunity.

Prompt: Act as a Chief Growth Officer with 30 years of experience, who has also successfully run captures for $100M contracts. From the solicitation notice and attachments, produce a two to three-page snapshot with evaluation emphasis, likely hot buttons, incumbent situation, teaming gaps, draft win themes or strengths (based on the attached documents on our company), and three shaping actions for this week. Tie each data point to a source quote or page.

4. Stakeholders and access plan: Once you are ready to proceed on a capture effort, begin documenting more specific capture action items and tasks for the team to execute.

Prompt: You are a senior capture executive having worked for midsized and large businesses in your 30 year career. Based on the opportunity capture snapshot and additional information provided (include other research you may have uncovered), list key stakeholders by role, involvement in the program, and focus area (i.e., influencing criteria) from a technical, program, or contracting perspective. Propose five purposeful questions for each stakeholder to test our solution approach and risks. Suggest the best next step to earn a meeting.

Verify Before You Act

• Cross-check against FPDS, USAspending, SAM, GovWin, and official agency documents.
• Maintain a living one-pager for each customer with dates and sources.
• Update the capture snapshot after every stakeholder touch so it reflects what you learned, not just what the model summarized.

 Gate Briefings

Question: What recommendations do you have for leveraging AI to support your gate briefing process?

Every company has a gate review process with executive briefings, checkpoints, and action items. Once your GenAI has enough context, it can help review and update your gate review process to ensure you’re capturing the data needed for informed decisions. AI can support the process, but judgment calls should remain human.

Using AI to Support Your Gate Briefing Process

The first step is to audit the gates themselves. Here is an example prompt:,

“You are an operations analyst auditing our capture gate process. Use only the documents I provide, do not infer strategy beyond the text. Review the process for completeness, clarity, and decision quality. Confirm that each gate has defined inputs, roles, decision criteria, and outputs that map to our corporate objectives. Return a concise report that includes: a findings table that quotes the source line and page, a per-gate checklist of required inputs, decision criteria, required approvals, and outputs, a simple roles and responsibilities matrix for who prepares, decides, and reviews, a Go/No-Go rubric with 6–8 scored criteria and suggested weights, and 4–6 process health metrics such as cycle time, hit rate by gate, aging pursuits, and PTW variance. Flag anything missing as “Not found” and propose short replacement language. Keep the review under 800 words and present tables where useful. End with three open questions we must answer to finalize the process, plus a one-slide outline I can drop into our briefing deck summarizing the top fixes and their decision impact.”

Based on your results, include these data points into your gate process:

Gate 1. Validation and early intelligence

Ask your GenAI to:

• Compile a customer snapshot. Include intended contract vehicle, set-aside implications, incumbent status and performance, current and future-state technical requirements, and potential evaluation criteria.
• Document obvious gaps that would require early teaming needs especially in cases where you may lack experience or connections.
• Produce a summary that highlights the identified requirements, evaluation emphasis, competitors, gaps, and immediate next actions. This could also be used for executive briefings.

Gate 2. Readiness and shaping

• Run a requirements analysis for the SOW and a draft compliance matrix tied to Sections L and M. If using GenAI versus a proposal AI tool, create these separately so that it accurately and comprehensively captures all of the requirements.
• Map your company’s past performance, experience, and key resumes to the requirement sections, so gaps become visible.
• Ask the GenAI to create win themes and translate them into evaluator-language strengths.
• Create a simple teaming decision matrix that shows what each partner brings.

Gate 3. Execution and risk

• Identify pursuit risks with probability, impact, owner, and mitigation actions.
• Use GenAI to workshop risks and document potential red flags that could affect capture success.
• Brainstorm mitigation strategies to address each scenario before it escalates.
• Assess early PTW levers such as labor mix, wrap-rate ranges, and fee structure to guide pricing strategy.

Briefing shell before every gate

GenAI can draft a briefing deck and weighted Go/No-Go view, but refine it using your own insight into the customer and program. Keep iterating inputs and outputs to strengthen each gate review.

 Up Next

In Part 3, AI in Action: Scaling Smarter Across the BD Lifecycle, we’ll explore how AI is changing capture and proposal team structures, how it supports cost and pricing strategy, and what’s next for fully integrated BD toolsets.

Get Up to Speed on the Full Series

 AI in Action: What’s Working in Federal Contracting? (Part 1 of 3)

AI in Action: Scaling Smarter Across the BD Lifecycle (Part 3 of 3)

If you want to move from experimenting with AI to operationalizing it across your growth lifecycle, Red Team can help. Our advisors can assess where AI fits into your BD, capture, proposal, and pricing processes, develop a compliant governance model, and build a practical roadmap your team can execute.