Monthly Recap: June '26

June was a big product month for us.

We kept Spring Shipping going — one demo video, one drop, every week — and published a new Hold The Churn episode we've been excited to share. Here's what went out.

What we shipped in June

Four releases from our 12 weeks of Spring ~~cleaning~~ Shipping series 🌸

1. Quivly MCP

Quivly now works inside the AI tools your team already uses.

Quivly MCP connects your workspace to Claude, ChatGPT, Gemini, and other chat surfaces — so customer questions get answered from live Quivly data, not a stale export.

→ Quivly MCP · Watch the demo

2. Ask Quivly

Ask Quivly is the conversation layer over your full customer graph.

Ask in plain English — "Which accounts are at risk right now?" — and get an answer grounded in health scores, usage, support, billing, and CRM data. It also pulls outward signals: funding, leadership changes, hiring, competitive moves. Inward and outward in one answer.

→ Ask Quivly · Watch the demo

3. Agent Builder

Agent Builder makes it straightforward to turn repeatable post-sales workflows into agents that scale across your book — predictable, auditable, and built for how your team actually works.

With AI, the best operators' playbooks become systems design problems. Agent Builder is how you encode them.

→ AI Agents on Quivly · Watch the demo

4. Data Models

Tanay walked through Data Models — the layer that makes everything else work.

This is how Quivly organizes customer data across integrations: core objects (customers, contacts, opportunities, subscriptions, tickets, usage), custom fields and objects, and field mappings that unify Salesforce, HubSpot, Stripe, your warehouse, and the rest into one graph.

Agents are only as good as the context they see. Data Models is that context — structured, mapped, and ready for AI.

→ Data Models docs · Watch the demo

More from Spring Shipping: YouTube playlist

Hold The Churn: Mike Kim, CEO of Mycroft

We published a new Hold The Churn episode with Mike Kim, co-founder and CEO of Mycroft.

Mike is building AI agents for security and compliance — a different domain from post-sales, but a very similar story: a lean team, no playbook, real customers, and agents doing work that used to eat human hours.

A few threads from the conversation:

• Scaling without a playbook — how Mycroft crossed 100+ customers by treating agents as force multipliers, not headcount replacements
• Sell before you build — Mike's take on validating demand early and letting customer pull shape the roadmap
• Full circle — Mycroft's post-sales team runs on Quivly while Mike builds agentic security on the other side of the house

→ Watch on YouTube · Spotify · Apple Podcasts · Podcast page

If this piques your interest, we'd love to talk — grab time here.