Visiting Engineer: Live Selling Copilot

Live commerce is bottlenecked by host capacity and operational chaos. Great sellers win on personality and community, but they still have to multitask like air-traffic controllers: managing chat, pacing listings, handling inventory, answering niche questions, and doing post-show follow-up. This role is to build a LiveOps Copilot: a human-first AI producer that runs backstage for live commerce sellers, keeps the host front and center, and automates the operational work around the show.

What You'll Build:
• A backstage producer cockpit that clusters chat, drafts or sends guardrailed replies, and surfaces the highest-value audience questions in real time.

• A listing and inventory control towebaer that helps hosts manage sequencing, pinned items, bundles, and oversell or stale-listing risk during the show.

• A real-time product research layer that answers unknown questions from seller notes, listing data, historical Q&A, and trusted category sources.

• Multilingual translation and localized reply support so hosts can engage cross-border audiences without losing flow.

• A lightweight post-show repurposing workflow that turns streams into reusable clips, Q&A moments, and listing assets.

What You'll Do:
• Build the end-to-end real-time system across chat ingestion, retrieval, guardrails, operator UX, listing controls, and post-show outputs.

• Work with one of our large LPs as your build/design partner to identify the right pilot category, host set, data path, and success metrics for the MVP.

• Design the automation ladder from draft support to host-controlled agentic replies without undermining authenticity or seller trust.

• Make decisions about latency, guardrails, disclosure, and escalation so higher-autonomy features are safe and useful.

• Work with AI Fund's build team on product wedge, technical risk, and the path from pilot to a broader live-commerce OS.

What You Need:
• Strong full-stack execution, especially on real-time systems, operator workflows, and AI-assisted products.

• Demonstrated experience building GenAI products with retrieval, tool use, guardrails, and production feedback loops.

• Product judgment for human-in-the-loop automation where the right answer is usually to remove busywork rather than replace the human outright.

• Ability to move quickly with external design partners and translate ambiguous feedback into working product decisions.

• You actively use AI coding assistants and modern GenAI tools as part of your build workflow.

• US work authorization. We are unable to sponsor visas for this role.

Helpful but Not Required:
• Experience with live commerce, creator tools, marketplaces, video products, or seller operations.

• Experience with multilingual UX, moderation systems, or low-latency chat infrastructure.

• Founder or founding engineer experience.

Who This Is For:
• A builder who likes real-time operator products, tight feedback loops, and markets where authenticity matters.

• Someone who sees the wedge in backstage tooling now, while building the data and control layer that could support deeper automation later.

What to Know Upfront:
• This is a 12-week, full-time, on-site residency in Mountain View, California.

• Not every residency becomes a company. The goal is to pressure-test the idea quickly and honestly with real users and customers.

• You will be building an AI Fund idea, not bringing your own startup idea into the program.

• The process typically includes a Builder Event or equivalent working conversation, then a 48-hour Build Challenge, then panel review with AI Fund build leadership.

• The compensation is intentionally modest during the residency because the upside, if the idea works, is a founder-level role.

Compensation:
$10,000/month for 12 weeks ($30,000 total). This is a contract role during the residency. If the build leads to a funded company, the next step is a founder-level role with meaningful equity upside.