Lightsage raises $4m to optimise software for AI agent buyers
Lightsage, a San Francisco startup founded by Jun Liang Lee and Sean Er, has raised $4 million in a seed round led by Nexus Venture Partners. The company is building what it calls an Agent-Led Growth (ALG) platform: infrastructure that helps software vendors understand how AI coding agents discover, evaluate and use their products, rather than optimising solely for human buyers.
The round includes backing from a range of operator-angels, among them former Salesforce CTO Steven Tamm, Postman CEO Abhinav Asthana, and Apollo CEO Matt Curl. Lightsage did not disclose its post-money valuation. Early customers named in the release include developer-tooling firms Firecrawl, Reducto, Daytona, Rime and Tinyfish.
What the platform does
The core proposition is that coding agents such as Claude Code, Codex, Cursor and GitHub Copilot can already select libraries, authenticate against APIs and integrate software products autonomously, without a human visiting a vendor's website. That breaks the assumptions underlying most modern growth analytics, which trace acquisition through browser sessions, clicks and sign-up forms.
Lightsage runs simulations across answer engines and coding agents, giving each agent real integration tasks and recording where the journey breaks down. Failure points may lie in discoverability, documentation clarity, authentication flows, SDK implementations or MCP server compatibility. Teams can patch the identified issue and rerun the simulation to confirm improvement. The platform also ingests real agent traffic data, attributing visits and product-usage events to specific agent sessions.
"We are moving from an internet where AI tells people which software to use to one where AI increasingly uses the software itself," said Jun Liang Lee, CEO and co-founder of Lightsage. "That changes what growth means. Visibility still matters, but the real test is whether an agent can understand your product and get to a successful outcome."
Market context
The ALG category is nascent but follows a well-worn pattern in developer software. Product-Led Growth (PLG) emerged as a discipline when self-serve SaaS made the end-user the effective buyer; Lightsage is positioning itself as the equivalent layer for autonomous software agents. The company draws an explicit parallel between Agent Experience and Developer Experience, the latter having become a recognised competitive differentiator over the past decade.
The closest adjacent category is Generative Engine Optimisation (GEO), a growing set of tools designed to improve a company's visibility inside AI-generated answers. Lightsage argues GEO addresses only the discovery step and stops short of the integration and usage stages where agent journeys most commonly fail. A number of well-funded startups are pursuing parts of this stack, and hyperscalers are beginning to offer agent-observability primitives within their own platforms, which will raise the competitive bar for pure-play entrants.
Nexus Venture Partners, which manages $3.2 billion in assets, has prior portfolio exposure to API-first and developer-tooling companies including Postman and Firecrawl, making Lightsage a logical adjacency for the firm. Partner Abhishek Sharma framed the investment as a bet on the shift of commercial agency from human to machine: the same optimisation industry that grew up around human digital behaviour will need to be rebuilt for autonomous agents.
What comes next
Lightsage intends to use the funding to deepen its evaluation, attribution and analytics capabilities across APIs, SDKs, CLIs, MCP servers and agent skills. Developer tooling is the initial vertical, but the company has signalled ambitions in B2B software, infrastructure and payments as agent-driven procurement broadens. Headcount growth across technical and commercial roles is underway.
The key near-term proof points for investors will be expansion beyond the current early-customer cohort, quantified improvements in agent-conversion rates, and evidence that ALG metrics correlate with measurable revenue outcomes for vendors. As AI coding agents grow in capability and adoption, the addressable market for agent-channel optimisation is likely to expand quickly, but the category remains at an early stage where no clear methodology standards have yet been established.