Why We Invested in Pryzm: Data-Driven Government Contracting

Last year, the U.S. federal government awarded roughly 11,000,000 contracts worth more than $700,000,000,000 for the goods and services that keep America running. 

Lockheed, Boeing, Raytheon - these are the companies we think of when we hear the term “government contractor.” And they are surely the most prominent, taking in billions each year with the Department of Defense as their primary customer. 

However, 95% of all government contracts are awarded to small and medium-sized businesses (SMBs) you’ve likely never heard of unless they are based down the street from your house. A whopping $291B was awarded by non-DoD Agencies too, soliciting everything from waste management to food services to office supplies. The US government already contracts with ~200,000 companies annually, but the economy supported by these contracts is so much larger—with thousands of subcontractors and service providers trying to get a piece of the already-awarded-pie. 

For anyone looking to get in on the public sector action here in America, the process is prohibitive to new entrants and the benefits of competition that come with them. Current business development workflows are intensely manual, with every contractor - even the longstanding ones - relying on brute force, expensive consultants, and often a little luck to stay in the game. Custom search engine software even exists for the troves of public data published by the government, but for one reason or another most contractors still resort to (1) repeatedly scrolling solicitation websites like SAM.gov on their own and (2) trying to qualify opportunities by cross-referencing solicitations with historical data pulled from award sites like usaspending.gov. 

This is first and foremost a time-consuming process that overwhelms contractors with noisy data, often causing opportunities to slip through unaddressed. When there may be a match, the act of qualification - or confirming an opportunity can be solved by what a contractor can offer by understanding the full context of an opportunity - is the real morale killer: Although the best contractors strive to anticipate the solicitation market (more on that later), the companies that are learning of an opportunity for the first time when it hits the open market have to rapidly process how to proceed:

When it comes to “contextualizing” the viability of government contracting opportunities for any business, the sequence today involves…

1. Pre-Qualification: First-Pass Relevance and Timing -  where teams must determine if it is in fact relevant and if they have time to respond. When using legacy tools in the market intelligence and capture space (like BGov, GovWin, and GovSpend), contractors still have to utilize lengthy manual keyword-based searches and comb through pages of irrelevant results.

2. Aligning capability offering to requirements and criteria of the government mission (and sometimes technology partners): this process is usually done by in-house program manager-type personas who know enough about both the business and product sides of their company to integrate both technical and sales teams in the context of what the government has asked for in solicitations. There is no platform for this workflow today, so it involves following the news, taking good notes, and going into government databases to find supporting material— which businesses currently hire consultants for. 

3. Identifying likely competitors and partners: Tools like GovTribe and GovWin can be used to collect data on companies, but it often involves endless searching, limited linkage between datapoints, and no extended workflow or platform for contextualizing that data to the specific proposal. 

On top of the noisy marketplace today, another serious contributor to this anachronistic process is how the government classifies solicitations—the NAICS Code framework (or North American Industry Classification System). To give you a sense of the limitations imposed by NAICS codes, there is a single code that categorizes all “software”––no matter how varied the products may be. Shield AI, for instance, has secured many federal contracts, but their market is hard to define with the current tools: when their Capture team searches SAM.gov for solicitations relevant to their AI Pilot software, the government database provides only 8 opportunities today, none of which are relevant. Other means of discovery include scattershot networking, and hiring expensive experts to help drudge up solicitations for companies to bid on but which are ultimately from the same pool. Within the current framework, it takes anywhere from 18 to 24 months to find, respond, and land federal contracts. 

Notably, for companies ‘in-the-know’, the process is slightly different. Contractors with previous knowledge, discussions, introductions, or program performance at the soliciting government office are already in the proposal development stage of the contract lifecycle well ahead of the competition and well before the public posting. They have the context already through non-public channels… often because they’ve helped create it in one way or another. 

Seeing that the Multi-Trillion-Dollar American public sector ultimately runs on the power of context, we are beyond excited to be backing Pryzm at first check. Their platform breaks down the opaque and convoluted world of government technology acquisition into something tangible, transparent, and intuitive so that any business can properly compete to give America the best solutions. While a slew of startups building in the space focus on proposal writing to increase the volume of proposals offered by a contractor, Pryzm understands that this is not the root problem preventing more companies from staying competitive in the Public Sector. The substantial challenge here is refining opportunity discovery to triangulate relevant information to the right opportunities, firing off fewer-but-better proposals, and then better anticipating the next ones. Pryzm leans into this problem set with a comprehensive and collaborative operating platform for growth teams to execute long-term successful federal business pursuits.

Pryzm was forged from Nick LaRovere’s experience at Palantir building federal operating systems and Matt Hawkins’ experience at Lockheed Martin managing some of the DoD’s biggest and most complex contracts. They teamed up with David Istrati given his expertise for complex data pipeline development at Veridion as well as a successful previous startup exit, and Justin Deckert on account of his operating experience with federal cybersecurity contractor Virsec. Then on top of it all, they were able to poach Raj Kane out of his AI/ML PhD program to firmly solidify the company’s engineering core. And the common link across this founding team? Maine’s liberal arts powerhouse (and emerging startup factory), Colby College - where they all first met and have remained tightly connected ever since. 

Zooming out, the market they’re going after is massive: federal contracts awarded in fiscal year 2023 reached a record high of $765 billion, reflecting a 9.5 percent growth from FY 2022. Although 200,000+ companies bid on government contracts in the US annually, the Pryzm team is sharp to point out that the market of potential customers expands far beyond that 200K figure, and for two reasons: “Pryzm’s data indicates that there are over 1 million unique organizations with distinct, active CAGE code registrations which, in our view, signals strong government contracting intent and capability from companies that simply may not have an active contract today.” Second, and more profoundly given their emphasis on integrating a broader set of context than their competitors, the ultimate market figure reveals a secondary market of more than 800,000 companies with the capacity to subcontract or provide service to the companies winning government contracts. 

This summer, the team closed on $2M from XYZ Venture Capital, Amplify.LA, and FirstIn, as well as several strategic angels, all of whom bring invaluable expertise in developing early-stage AI companies at the intersection of technology and the national industrial base. 

If you learned something new or want to support the Pryzm mission alongside Amplify, consider giving this a share and passing it along to anyone else who you think may find this relevant.

Amplify is a first-check fund backing LA’s earliest and strongest tech teams. Want these updates delivered to your inbox? Join our newsletter!