Fulfilling government contracts requires procurement professionals to spend weeks methodically combing through databases and documents, all against tight deadlines.
Now, this is rapidly changing as artificial intelligence transforms this traditionally manual process and provides new tools that can reduce proposal writing time by up to 70%.
Joe Schulman, partner and principal at PricewaterhouseCoopers (PwC), who leads the firm’s AI services for the aerospace and defense sector and co-leads the firm’s U.S. space program, believes that the adoption of AI will help companies He said the way government bids are developed is changing.
He says AI is streamlining nearly every aspect of the government contracting process, from identifying opportunities to drafting complex technical responses.
In April 2023, PwC announced a $1 billion investment to grow and scale its global AI services. The company has created specific AI solutions for aerospace and defense, including Flight Science and Engineering Design AI and PartsGPT. “However, AI for proposal writing is the most frequently requested application in our practice,” Schulman told SpaceNews.
These are not your typical chatbots. A new generation of AI tools is specifically designed to navigate the complex world of government contracting, where security protocols are stringent and risks are high, where a single proposal can make or break a company’s fiscal year.
Beyond basic automation
This technology goes far beyond just document processing. Modern AI systems can scan the vast database of government requests on SAM.gov (the official U.S. government procurement portal) to identify contracts that match a company’s capabilities. We can also help you break down complex requests for proposals (RFPs) into manageable tasks and draft proposals that highlight your company’s strengths. At the same time, we ensure compliance with strict government regulations.
For example, for niche contractors in the space industry, the ability of AI to precisely target relevant opportunities eliminates the need for countless hours of manual searching of federal databases, Schulman says. “For commercial space companies, it can be difficult to determine which Space Force contracts to bid on,” he explained. “This technology helps you avoid wasting time chasing mismatched opportunities and focus specifically on the sweet spot.”
Reforming the proposal process
The traditional response process for government RFPs has long been resource-intensive. Teams of writers, subject matter experts, and compliance personnel typically spend weeks parsing hundreds of pages of requirements to ensure that every specification is met and every question is answered. Check. Some RFPs have deadlines as short as 30 days, requiring companies to submit comprehensive proposals quickly.
“We are now saving at least 60 to 70 percent of time by removing these remediation tasks and leveraging these tools as accelerators to support our response,” Schulman said. AI systems can automatically summarize the entire RFP, break it down into actionable items, and automate responses to proposals.
One particularly powerful application is the ability to process large volumes of government regulatory documents. For example, in discussions with satellite providers, PwC provided a proof of concept that leveraged generative AI to better understand NASA certification requirements, Schulman said. He said it also helped him prioritize which requirements needed to be completed first.
The human element remains important
Despite the increased efficiency, industry experts stress that AI will not replace human expertise. “Our stance as a company is that AI is not a replacement tool. It’s an accelerator tool,” Schulman said. “We want individuals and experts within our company with that level of intelligence and passion for a project to inject some human-based creativity into the answers to our proposals, but with the help of generative AI services, , removing a potential human step from the process.”
This view is echoed by Sean Williams, co-founder and CEO of AutogenAI, a London-based startup that is expanding its AI-powered proposal writing services in the US.
“Humans are much better at understanding logical context and the nuances of an argument,” Williams said in an interview. “AI generates the ideas, but humans craft the responses.”
security compliance
One of the biggest challenges when implementing AI in government contracting is ensuring data security compliance. Commercial AI tools such as ChatGPT are not allowed in secure government cloud environments required by defense contractors. Therefore, AI tools must be built on approved platforms.
Schulman said PwC’s solution leverages Microsoft Azure OpenAI. It is one of the platforms approved to process DoD proposal data in compliance with the strict security standards of the Department of Defense and the International Traffic in Arms Regulations (ITAR).
Amazon Web Services, Google Cloud, and Oracle Cloud were also recently cleared to host sensitive Department of Defense workloads.
Cloud systems must meet Department of Defense security requirements known as “impact levels.” This requirement ranges from Level 2 for public information to Level 5 for sensitive mission-critical data. “It’s great that these capabilities are finally operational at DoD impact levels 4 and 5 and even classified levels,” Schulman said. “We’re going to see great progress.”
Demand for AI-powered proposal writing and production tools has attracted some venture investment. AutogenAI’s recent $39.5 million Series B funding round co-led by Salesforce Ventures and Spark Capital brings total venture investment to $65.3 million.
Nowi Kallen, managing director of AI at Salesforce Ventures, says it’s becoming clear that this technology is poised to transform time-consuming and highly complex bidding processes.
military perspective
The military itself is also cautiously introducing AI into its procurement processes. Col. Heather Bogustie, who oversees space systems integration for the Space Force Space Systems Command, said the Department of the Air Force recently deployed NIPR-GPT, a generative AI program for use in the Unclassified Internet Protocol Router (NIPR). Point out. Secure military network.
“This will make the use of AI more prevalent and championed, and remove a lot of the stigma,” Bogusti said at the recent Milsat Symposium in Mountain View, California.
The military eventually aims to use AI to evaluate industry proposals, but this is currently a completely manual process. “We want to be able to leverage this even further in the procurement lifecycle to help with things like source selection, technical evaluation, proposals, and RFPs to speed up the day-to-day operations of the business.” Bogusti he added.
AI learns government dictionary
Some companies are already realizing the benefits of implementing AI. Aalyria, a space communications technology company spun out of Google’s parent company Alphabet, uses large-scale language models to tailor proposals to specific institutions’ technical language.
“In our industry, we often have as many as 10 different terminologies to describe the same thing,” explains Brian Barritt, Chief Technology Officer at Aalyria. “We found that training the AI based on past texts and having it respond in the appropriate language that the government expects can be very helpful in supporting responses to suggestions.”
The company developed a system that can be trained on previously created proposals and agency-specific terminology to ensure new submissions match the expected language and format. This attention to language detail is critical in government contracts, where precise terminology can make or break a bid.
Future prospects and challenges
AutogenAI’s Williams said there is still no concrete data on whether AI-powered proposals will win more contracts. However, early feedback suggests the technology is paying off. “It will probably take another 12 months for government procurement to be completed, and then we will have hard evidence. But early signs and early anecdotal data are very positive.” said.
Technology continues to evolve rapidly. But PwC’s Schulman emphasizes that human oversight remains important.
He explained that AI hallucinations are common. “Therefore, there will always be a need for a human to handshake with the AI to meet the required level of quality and compliance. The relationship between AI and humans, especially in aerospace and defense, is delicate. Similarly, the balance of responsibility changes. What we want to avoid is one-sidedness.”
As more companies adopt these tools and the technology matures, he predicts the government contracting landscape could change significantly in the future. While AI can accelerate and enhance the proposal process, human judgment remains essential in the business of government contracting.
This article first appeared in the November 2024 issue of SpaceNews Magazine.
