Locke pairs AI agents with lobbyists to chase government contracts

Why it matters

Locke is testing whether vertical AI can improve the margins of relationship-heavy professional services. Its $500K run-rate claim shows early demand; the human workload will determine whether the model scales.

Parth Badhwar (@parth_badhwar) introduced Locke in early August as an AI-powered government affairs operation that combines software agents with in-house lobbyists to help companies monitor regulation, influence policy and pursue government contracts.

Locke said in its Y Combinator launch post that it exceeded a $500K annual revenue run-rate within six weeks of quietly starting operations. The figure is self-reported and represents an annualized pace, roughly $41,700 a month if sustained, rather than a full year of recognized revenue. Locke described its customers only by sector, citing defense, healthcare and government technology.

The hybrid structure is central to Badhwar's pitch. Locke's agents analyze legislation, monitor stakeholder engagement and draft contract proposals. Its human staff handle meetings, lobbying and the relationship work required to move a proposal from research to a funded contract.

Badhwar arrived at that model through government and sales rather than an engineering lab. According to his public biography, he worked in the Biden-Harris White House's Office of National Drug Control Policy, on Deloitte's policy and government relations team, and as a battleground states associate for the Harris presidential campaign. He earned a political science degree from George Washington University.

Immediately before founding Locke, Badhwar led go-to-market work at Clipbook, an AI research platform for communications teams. Y Combinator says Clipbook sold to dozens of government offices, giving Badhwar experience selling software into the institutions Locke now wants to help its clients navigate.

Software wrapped around a services firm

Locke is applying the vertical AI playbook to a market where access, judgment and personal relationships remain part of the product. Rather than selling regulation-tracking software and leaving customers to hire outside advisers, Locke plans to own the work from monitoring and compliance through lobbying and procurement.

The Locke Government Affairs service tracks legislation, regulations, hearings and requests for proposals, with alerts delivered through Slack. Locke also offers compliance briefs, stakeholder mapping, proposal drafting and lobbying at the state and federal levels. Locke says its target sectors include defense, robotics, healthcare, energy, space, advanced manufacturing and critical infrastructure.

That breadth makes Locke closer to a software-enabled professional services firm than a conventional software subscription business. The agents can reduce research and drafting time, while lobbyists and policy staff remain responsible for outcomes that depend on agency priorities, legislative calendars and human relationships.

Badhwar has explicitly rejected the idea that software alone can remove those constraints. "You cannot simply deploy software around them," he wrote in an August 5th essay about government processes and the people who administer them. His strategy is to use AI to compete inside the existing system before attempting to replace parts of it.

Locke's initial hiring plans reflect the split. Its Y Combinator profile listed openings for founding technical and policy staff, with the technical role based in San Francisco and the policy position spanning San Francisco and Washington, D.C. Locke was founded in 2026 and joined Y Combinator's Summer 2026 batch.

The $10 trillion pitch

Badhwar frames the opportunity around more than $10 trillion in annual spending across federal, state and local governments. That total is a market narrative, rather than an estimate of revenue available to Locke or its customers. Government spending includes transfers, salaries, existing programs and other expenditures that will never become competitively awarded vendor contracts.

Locke's practical market is narrower: companies willing to pay for regulatory intelligence, lobbying and procurement support. That remains a substantial services category, particularly for startups in regulated industries that lack their own Washington operations.

The early run-rate claim offers evidence that some customers will buy the combined service. It says less about the economics Locke can produce at scale. Lobbying and contracting work can be labor-intensive and tied to long government cycles, making automation rates, client concentration and staff utilization central to the model.

Locke's bet is that agents can compress enough of the research, monitoring and document work to let a small policy team serve more clients than a traditional government affairs shop. The lobbyists still have to get through the front door.