From AI Assistance to AI Orchestration: Key Takeaways From Outreach's Explore AI Webinar

Revenue teams are stuck between competing demands: grow pipeline faster while cutting costs, close more deals with fewer resources, and somehow make every interaction feel personalized at scale. If that sounds impossible, you're not alone. 

I had the opportunity to moderate our recent Explore AI webinar including a detailed conversation with Nithya Lakshmanan, SVP of Product Management about how Outreach is solving this exact problem, and many other seller pains, with agentic AI. Brian Goldfarb, CMO of SolarWinds, also joined the webinar and gave us an insider’s view sharing real results from his team's agentic AI journey with Outreach. 

If you haven’t had a chance to watch the webinar yet, here's what you need to know. 

The shift from systems of record to systems of action 

We're at an inflection point in sales tech. For years, CRMs served as digital filing cabinets — useful for storing data, but not for actually getting work done. Revenue orchestration platforms like Outreach are changing that equation entirely. 

Instead of jumping between your CRM, data provider, LinkedIn, and conversation intelligence tool, everything now lives in one place. Generative AI pulls from your customer data, engagement history, and public sources to give sellers full context before they even open an email. And with agentic workflows, AI can reason across these systems in real time, surfacing patterns and risks that humans simply can't catch at scale. 

The old playbook was data → insights → action. The new playbook is AI that acts. Every seller moves faster and smarter, taking the right step at the right time. 

Why AI adoption is still failing (and how to fix it) 

Despite all the hype, most AI projects aren't working. MIT research shows that 95% of generative AI pilots fail, and Gartner predicts over 40% of agentic AI projects will be canceled by 2027. 

The problem isn't the technology. It's how companies are implementing it. 

Nithya identified three major challenges blocking AI adoption: 

• Unclear use cases. Teams deploy AI without defining what problem they're solving or how they'll measure success. The tools get rolled out, but nothing actually changes. 
• Tool overload. Sellers are already overwhelmed. They don't want to learn another platform or chase down another dashboard. AI should work like a teammate — helping behind the scenes, nudging at the right moment, not adding to the workload. 
• In-house builds that lack domain depth. Many companies try to build agents internally for control and customization, but these tools miss the guardrails that make AI trustworthy. The same MIT report found that AI projects with external vendors have twice the success rate as internal builds. 

And then there's agent washing. Too many vendors slap "AI agent" on basic automation or chatbots, which creates confusion and kills momentum. 

H2: How Outreach approaches AI differently 

Outreach doesn't just claim to be AI-powered. We've built AI that's simple, relevant, and trusted — designed for how revenue teams actually work. 

• Simple means AI is built into the seller's natural workflow. The goal isn't to turn every rep into a prompt engineer. It's to help them feel more confident and in control while AI does the heavy lifting in the background. 
• Relevant means AI has access to the right customer data and your company's internal know-how. Every recommendation and action is spot-on because it's pulling from the sources that actually matter. 
• Trusted means every AI action is governed, explainable, and auditable. Sellers stay in the loop for critical workflows. You can't just hand everything over to AI and hope for the best. 

Outreach was the first to market with ISO 42001, the global standard for responsible AI governance. But we go beyond compliance. We don't just use AI to summarize or surface insights — that's table stakes. We use AI to reason across all your interactions, spot risks, and orchestrate actions that actually progress deals. 

Here's an example: if a new decision maker joins one of your target accounts, Outreach automatically identifies that person, scans previous conversations with that company, figures out what's relevant to their role, and engages them with tailored messaging to bring them into the deal. That's not AI taking notes. It's AI driving outcomes. 

What's new in Outreach's Q4 release 

Outreach shipped a wave of new AI agents and platform improvements designed to make sellers more productive and sales leaders more confident in their forecast. 

New AI agents available today 

• Research Agent is your digital researcher. It automatically gathers insights from public data — websites, 10-Ks, press releases — and your conversation data. It surfaces customer pain points, needs, and even sentiment toward competitors. You can pinpoint companies investing in AI, expanding internationally, or signaling growth. The Research Agent also updates account plans and streamlines handoffs when accounts change hands between AEs or from AEs to CSMs.
• Personalization Agent takes all that research plus first-party and third-party data and turns it into messaging that feels human. It works across channels — email, call scripts, LinkedIn messages. These messages can be sent autonomously or handed off to sellers in their inbox for a quick review and tweak. Either way, it saves reps 80% of the time they'd spend manually writing personalized communications. 
• Revenue Agent orchestrates entire sales plays end to end. It finds the right prospects, enriches them with third-party data, researches them, and enrolls them into AI-personalized journeys — all in one workflow. You can spin up AI SDR agents for top-of-funnel outbound or inbound, and event agents that automate the promotion and follow up of your events. 

• Deal Agent picks up right after a meeting. Instead of manually updating your CRM, the Deal Agent does it for you. It automatically logs qualification criteria like new champions, competitor mentions, and next steps. 
• Call Agent creates personalized call scripts and drops personalized voicemails. Sellers don't waste time trying to connect with prospects who aren't picking up — the agent handles leaving a voicemail on their behalf.

Platform improvements that drive efficiency

Beyond agents, the Q4 release includes features that make AI feel like part of the daily workflow, not an added burden. 

Personalized homepage for different SDR, AE or CSM workflows that shows customized information like your top opportunities, accounts, and upcoming meetings. 

Team outcome report uses AI to help managers understand what activities and conversation topics impact win rate, so they can coach teams to adopt best practices. 

Expanded data enrichment with new signals from  ZoomInfo and SalesIntel. 

• ZoomInfo intent data uncovers what key accounts are researching so reps focus on accounts that are in-market for your product. 

• New signals from SalesIntel allow Outreach AI agents to capture IPO and acquisition activities to create more precise account plans and relevant messages. 

What's next?: The future of conversational AI – Ask Outreach

The next wave of innovation is already taking shape. Outreach is developing a Model Context Protocol (MCP) that lets Outreach AI work seamlessly with other agents your customers already use - so sellers can deliver smarter, faster, and more connected experiences without switching platforms. Additionally, a new conversational experience called Ask Outreach will sit center stage in your personalized homepage. 

Instead of clicking through dashboards or hunting for reports, you can simply ask questions like "show me my top deals this quarter" and then follow up with "which of these are at risk?" The system surfaces answers and suggests the next action in one connected flow. 

The conversation thread continues with context, so you can pick up where you left off even across sessions. And you can access Ask Outreach anywhere you work — Slack, Microsoft Teams, or the mobile app. 

Outreach is also expanding its family of agents with: 

• Reply Agent taps into your internal knowledge sources — enablement content, pricing FAQs — and automatically generates responses when a customer asks about your product, pricing, or a competitor. Sellers don't have to hunt for information or track down experts. AI produces the best response so the deal keeps moving forward. 

• Next Best Action Agent looks across all your engagement signals — who's opening emails, which prospects are active, who's gone silent — and translates that into clear actions. You'll never miss a follow-up or overlook a stakeholder. It also uses buying intent from market data like when a company is hiring or expanding and turns that knowledge into action.  

• Meeting Agent preps you before you even join a call. It briefs you on buyer context, recent interactions, talking points, recommended content, and every attendee joining the call. Afterwards, it tracks all action items so nothing slips through the cracks. 

Real results from SolarWinds 

The insights and visionary focus for Outreach is quite compelling, but what really brings it to life is hearing from one of our early AI Agent customers, Brian Goldfarb, CMO at SolarWinds. Brian was very kind to share how his team uses Outreach to pilot agentic AI use cases and drive measurable results. 

SolarWinds started with email personalization. When they compared human-written emails to AI-generated emails, Outreach’s personalization agent won every time. AI creates better content because it does the work — the research, the signal gathering — and uses first-party and third-party data to craft messages that actually help you connect with your customers. 

The results speak for themselves. SolarWinds is seeing better open rates, better reply rates, and fewer unsubscribes. In some cases, they're seeing 300% better metrics. And it's not just leading indicators. They're also seeing meetings booked and opportunities created — actual pipeline generation. 

SolarWinds is also using conversation intelligence (KAIA™) to capture conversation data and feed it into their broader AI workflows. They're running renewal motions where all calls are recorded, transcripts are extracted, and data is fed into AI models to identify what actually drives deal outcomes. They're determining the metrics that matter — if a call lasts this long, what's the probability it closes? If there's a two-way dialogue for a certain percentage of the call, is that an indicator? 

But Brian was also transparent about the challenges. Governance is the thing that scares him every day. They've had agents react unpredictably so they're building guardrails and keeping humans in the loop. They're also tackling data management — making sure the "golden corpus" of data feeding their AI models is clean, up-to-date, and relevant. 

Looking ahead, Brian's goal for the next year is to get humans out of the loop on certain workflows, figure out the financial operations of AI (because tokens aren't cheap), and move toward a future where agents talk to other agents to orchestrate work autonomously.

The path to self-driving revenue teams 

Outreach's vision for the future is clear: a self-driving revenue platform where AI becomes your teammate on every task, in every job. 

It starts with a context layer — a unified intelligence foundation that connects signals across CRM, conversations, first-party and third-party platforms, and even internal knowledge. This gives AI the depth it needs to act with precision. 

On top of that, Outreach is building agentic workflows that fall into three categories:

• Self-updating agents automatically update systems on behalf of sellers. AI keeps you on top of deals and accounts, automates follow-ups, and handles most of the admin work. 
• Self-guiding agents coach sellers in the moment. They surface risk, opportunities, and provide coaching so sellers can hit the best outcomes.

• Self-learning agents get smarter over time. They learn from past interactions and outcomes, rebuild their own memory about you as a seller, and automatically adjust strategy to get the best results.

And to be clear: self-driving doesn't mean no sellers. Sellers are still doing the critical work of relationship building and closing deals. But now every seller operates at the level of your best seller because AI is providing updates, coaching, and learning alongside them. 

The takeaway 

The teams winning in 2025/2026 are the ones that consolidate tech stacks, coach with AI in the workflow, and extend AI across the entire revenue cycle — not just prospecting. These aren't small changes. They're foundational shifts that determine who keeps up and who falls behind.