As an account executive at Airbyte, I'm working multiple deals at any given time. Each one involves multiple stakeholders: engineers, legal, security, a champion who's selling for me internally, and a decision-maker I may or may not have met yet. Some of these deals take over four months to close.
So I spend a real chunk of every week just getting caught up. Checking Gong to remember what a customer said on the last couple of calls. Checking my calendar for what's coming up. Scanning Slack threads with our solutions engineers for their notes. Before I can actually move a deal forward, I have to piece together its current state from five different places.
Without a way to organize this context, you start relying on memory. That's where things get risky. You forget a blocker, or a concern legal brought up. And in enterprise sales, where deals are multi-layed and complex, that can be fatal.
I built Mission Control to take care of that for me.
What It Actually Is Mission Control uses Airbyte's MCP to pull data from the systems I work in every day into one place. Right now it connects to Gong and a few other tools. When I open it, I see the deals I'm tracking for the week, the action items I need to take care of, and the context from my last few customer conversations.
From there I can see every open loop across my pipeline. I know who owns the next move, and when we last talked.
Each account has its own deal card. What’s inside the deal card? Everything I need to prepare: customer quotes, a stakeholder map, open blockers, key gaps in the deal, and the next planned actions. Pulling the right prep materials together for each account is what a sales collateral finder agent does ahead of a call.
All of this data already exists. It's just spread across a bunch of tools that don't talk to each other. Airbyte's MCP connects them, so now I have one screen instead of fifteen tabs.
Why I Built It Myself I work at a company that makes data infrastructure. We have the connectors and we built our own MCP. Solving my own problems with our product makes it a lot easier to sell it to customers.
The build was pretty easy, actually. I connected Airbyte's MCP to Gong, Calendar, and the other tools I use, and it just worked. I logged into each app, told it what data I wanted, and started building the view on top. Most of my time went into figuring out what I actually wanted to see, not getting the tools to connect.
Why Airbyte's MCP? A lot of the tools I use have their own plugins or community MCPs available, so it's not like Airbyte's MCP is the only way to connect them. But the value is that I can hook up one MCP and connect everything through there. Way simpler. And I know I can trust the data coming through. I'm biased, but I trust a data movement platform's connectors and infrastructure more than a community MCP or a straight API plugin.
So the only thing I actually had to figure out was what I wanted the dashboard to show me. I used Claude to build the interface, and that part went fast because I already knew what I needed: which deals need attention today, what happened on the last call, what I promised the customer I'd do.
What It Changed I used to spend about 15 minutes before every call just gathering context, flipping between Gong, Calendar, Slack, and my CRM. With Mission Control, the context is already there. I'm down to a few minutes of prep now.
It also changed how I spot risk in my deals. When everything is on one screen, you notice things. A key stakeholder who hasn't shown up in two weeks. A technical question that's been sitting unanswered. Those signals were always there in the data, I just couldn't see them when they were spread across different tools.
Where It’s Going I'm still adding to it. Gmail and calendar connectors are next so I can see email threads alongside my call notes. I also want to connect it to Airbyte's Context Store . The idea is that instead of the agent searching across every tool every time I ask it something, the Context Store already has the relevant data ready to go. That means faster answers and better ones, because the agent isn't sorting through noise to find what matters. For someone who's asking it questions between calls all day, that difference adds up.
Look, I'm an AE, not an engineer. Building software isn't really in my job description. But I connected Airbyte's MCP to the tools I already use, and a few days later I had something that actually makes my week easier. If a salesperson can build on top of this, I think that tells you something about how accessible it is. The same accessibility is what lets go-to-market teams build ai agents for marketing on top of the data they already have, without waiting on an engineering sprint.
