Scott Schlegel

🔈 Today, I'm very excited to announce the release of Ashby's built-in candidate fraud protection system. Candidate Fraud has been a large and growing problem for TA teams - especially those hiring remotely for technical roles. At Ashby, we care deeply about being responsive to the changing needs of our customers. As this trend started to emerge, we pulled it forward on our roadmap. I'm very proud of the approach the product & engineering team took in building this functionality. They moved with urgency, but also spent the last few months working with design partners to iron out this functionality, resulting in a thoughtful solution to the problem. When this feature is enabled, we scan inbound applications for dozens of signals across a few core categories (device anomalies, location mismatches, suspicious online identities, etc.). We then transparently flag any anomalies to the TA team for review. This approach means that legitimate candidates aren't required to take any extra steps to apply, and it means that each team can make its own risk determination of how they want to treat different signals. The early impact we've seen with design partners has been very significant - teams had been interviewing a non-trivial number of fake candidates and were able to bring that number to zero. This functionality is live for all Ashby customers today! Scanning inbound applications for these signals is quite expensive, but we wanted to make sure we don't gate this functionality for any of our plans. As a result, the feature utilizes our existing AI credit system and consumes 2 credits per scanned inbound application. Since we recently increased the credits included in each plan, the vast majority of Ashby customers have a large pool remaining and can turn this feature on right away (you can enable it globally, or turn it on job by job). For folks making heavy use of this feature and requiring more credits, we have an exciting bundle we're launching as part of our AI launch event on 9/25! I'm excited about the impact this will have on both improving the security practices of TA teams out there and giving valuable time back to core hiring activities. This is a busy launch month for us, so we'll back with more updates soon!

457

51 Comments

We spent months trying to sell debt repayment companies on debit card rails. Zero traction. Complete waste of time. Then we interviewed target customers more deeply. Turns out debt repayment companies didn't need new payment rails. They needed ACH that actually worked for their specific workflows. Custom scheduling. Exception reporting. Data elements that matched their existing systems. So we rebuilt our ACH processing from the ground up. Made it work exactly how debt repayment operations needed it to work. The volume doubled on ACH alone. Only then did those same companies start asking about modern payment options. Once we'd solved their real operational problems, they trusted us with the new stuff. The pattern holds across every vertical we've entered. Fix the boring, critical stuff first. Make their existing processes work better. Earn the right to introduce innovation.

194

27 Comments

I’m back this week with Tech 5 - this time with Andy McKinney of Able. Andy and his team are doing remarkable work with AI, and we’ve been fortunate to collaborate with them on the next big things in software development at Riskonnect, Inc.. What stood out most in our conversation wasn’t just the technology, but how critical people and culture are to success.  Andy made a powerful statement - “I think there have been times when Agentic AI has been put forward as a solution looking for a problem. If you’re starting with the technology, you may wind up solving a problem that’s the wrong one to solve.” I agree: tech transformation isn’t about the tools alone. It starts with strategy — then people, process, and technology enable it. The 3 points that stuck with me: -          Strategy first, technology second: Outline your business goals first and then implement the tech to achieve them. -          Context before agents: Just as setting clear end objectives and success metrics for a new employee entering the system is important, this also applies to introducing agents, the latter being much more complex. -          Overcoming resistance to AI: Getting teams hands-on with AI at boot camps will help to change their mindset and get them to be more open to working with it. At the end of the day, AI should help us grow and innovate while keeping people at the center. How have you prepared your teams for AI adoption — culture first, tech second? #AgenticAI, #Riskonnect, #RiskManagement, #TechnologyRisk, #ProductDevelopment, #GenAI 

41

3 Comments

We’ve made the hard part of agent infrastructure… slightly less hard. Today, we launched something I’ve wanted to exist for a long time: managed execution and credential governance for AI agents. It’s not glamorous. It’s not flashy. But it is necessary. Most agents today are a duct-taped mess of local code, hardcoded secrets, half-finished OpenAPI specs, and demo workflows that fall apart under real-world pressure. We’ve all seen it. This release is our first proper step toward fixing that. Any agent can now execute APIs and workflows through Jentic’s hosted infrastructure, with real security and full control. No local secrets. No prompt injections. Just a better way to run agents. That means: * A secure credential vault * Agent-level permissions * Execution over MCP or REST * A catalog of 1,500 APIs and 2,000 workflows * A new web app to manage it all We’ve flagged some integrations as “Jentic Tested” where we’ve done the hard work ourselves. The rest? Still wild. Still worth taming. And I hope you’ll help. We’re also opening the waitlist for our hosted Jentic agent. If you want to test the full stack with no setup, sign up at the link in the first comment. Let’s build the layer agents actually need. Jentic Sean Blanchfield Dorothy Creaven

63

7 Comments

When Kevin Coleman and I started Ravenna, one of the first things we did was sit down with Ops leaders and ask one simple question: where does your team actually work❓ The answer wasn’t a portal. It wasn’t some legacy dashboard with its own login, training module, and adoption curve... It was Slack. Teams. In Slack and Teams, HR managers answer questions about benefits and PTO. IT directors reset passwords, unlock accounts, and ship out new laptops. Operations teams track down approvals, chase missing data, and route requests across departments. That’s where the real work piles up, and where it needs to be automated. The incumbents technically “integrate” with chat, but the pattern is backwards. Their apps use Slack as a notification layer, just to push you back into their portal. That anti-pattern is what kills adoption. People don’t want to leave the environment where they’re already working. The fix is obvious: keep the work in chat, not in a portal. Ravenna lives entirely inside Slack/Teams. The bot handles requests in-line, so the problem is solved where the conversation starts. Need a new laptop? Submit the request in Slack, and it’s approved and ordered without leaving the thread. Locked out of Salesforce? Ask in Teams, and the bot helps you restore access quickly. Looking for the PTO policy? The exact page from the handbook surfaces right where you asked. Ravenna is designed for convenience on purpose, but the real outcome is adoption. Drop a separate portal in front of employees, and usage falls off a cliff. Keep the workflow in Slack or Teams, and people actually use it. That’s where you capture the signals: who’s asking for new laptops, who needs Salesforce access, who’s stuck on an HR policy. And once you capture those signals, you can automate the repetitive requests that eat up time. This is the philosophy we’re betting on: co-location beats portals every time 🚀

34

1 Comment