Taking the Sexy Out of AI: AI for Ops

Artificial intelligence is no longer just a buzzword - it’s a tool being pushed heavily by executives and hyped across LinkedIn. But for RevOps professionals, the real question is: what actually works, and how do you avoid spinning your wheels chasing AI hype?

‍

That’s exactly what Camela Thompson (Head of Marketing at RevOps Co-op), Courtney Sylvester (Director of RevOps at Openprise), and Dom Freschi (Customer Success Director at Openprise) tackled in this practical, grounded discussion about implementing AI for operations.

‍

The FOMO Is Real - But So Are the Challenges

‍

AI is everywhere. But just because you can use it doesn’t mean you should.

• Leaders are pressuring Ops to "do something with AI," often without understanding the foundational work required.
• AI implementation is rarely plug-and-play. It demands technical knowledge, data quality, and iterative testing.
• Ops teams often inherit these tools without additional headcount, making scalability difficult.

‍

“Everybody wants to try it, master it, play with it—but without a strategy, you’re just scaling bad practices.” – Camela Thompson

‍

AI Works Best When It Has Clear Boundaries

‍

Rather than trying to automate everything, focus on areas where AI can support repetitive, data-heavy tasks:

• Use LLMs to standardize and enrich job title data for segmentation.
• Summarize email threads, Gong calls, or Salesforce notes to support better handoffs and sales context.
• Scrape external data - like security breach news - for trigger events and route to sellers.

‍

However, AI is built to sound human - not to be right. That means hallucinations and incorrect outputs are inevitable without tight guardrails.

‍

Start Small and Build with Purpose

‍

Courtney and Dom emphasized the importance of beginning with specific, high-impact problems. This allows teams to:

• Learn prompt engineering and input/output patterns.
• Set up proper QA workflows to catch errors early.
• Avoid building tools or agents that work well once but break at scale.

‍

“Don’t just build something and walk away. Every AI process needs checks, balances and constant monitoring.” – Courtney Sylvester

‍

AI Will Amplify Your Data Quality - for Better or Worse

‍

Your results are only as good as your data. The panel warned against using AI when your CRM or enrichment data is messy:

• Bad input = bad output.
• AI is not a fix for poor hygiene - it’s a magnifier of it.

‍

To mitigate this, Openprise uses AI to flag anomalies, validate LinkedIn URLs, and identify duplicate or stale records - especially when mapping contacts across accounts.

‍

For more on how RevOps can and should build with AI, check out RevOps, now you can build with AI.

‍

More Technical = More Valuable

‍

The future of RevOps is more technical, not less. Understanding system architecture, APIs, and even programming logic (like Python or SQL) allows Ops to:

• Build more scalable AI workflows.
• Directly integrate with OpenAI or custom LLMs.
• Reduce reliance on expensive middle-layer vendors.

‍

“I don’t think it’s necessary for everyone to learn code. But for those who want to, it opens a lot of doors.” – Courtney Sylvester

‍

Workflow Examples You Can Steal Today

‍

Want to see where AI is already driving ROI? Here are some of the real-world AI workflows Courtney and Camela are running:

• Summarizing opportunity context from historical activity logs.
• Scraping job boards for hiring intent and pairing with news scraping.
• Parsing LinkedIn data to validate contact matching across records.
• Identifying account anomalies (like broken websites or acquisitions).
• Using AI in lead scoring - based on scraped insights, not gut feel.

‍

Just be careful: deterministic logic should still drive core workflows like lead-to-account matching and routing. AI is not ready to make those calls alone.

‍

For more on building AI agents, check out Build custom AI Agents for RevOps: introducing AI-Agent Factory.

‍

Final Thoughts: AI is Just One Tool in the Ops Toolbox

‍

The temptation to chase shiny tools is strong - but AI is not a replacement for understanding your business, knowing your data, and building for scale.

‍

“Instead of rushing into wild workflows, start with the problems you’ve always had but never had time to solve.” – Dom Freschi

‍

Also check out This isn’t the AI we were promised – session recap from RevOpsAF to hear more about Courtney and Dom’s POV on this topic.

‍

Looking for more great content? Check out our blog and join the community.

‍