Building with AI: A Product Ops Community Talk

Agents, automation, vibe coding — product ops teams are building with AI in ways that didn’t exist a year ago. But what does that actually look like inside a real product ops team?

In this community-led session, Product Ops HQ members share what they’ve been building, experimenting with, and learning across their workflows:

🎤 Pourush Kalra shares Meeting Maestro, the tool he built to turn meeting transcripts directly into Jira stories, cutting hours of manual write-up work.

🎤 Neha Tiwari from Sayari walks through her team’s end-to-end intake-to-execution workflow — from Salesforce intake, through triage, to R&D planning and prioritization — plus how her team uses MCP to agentically break initiatives into epics and stories, and how Claude powers their scheduled exec reporting straight from Dragonboat data.

Whether you’re deep into agents or just getting started, this is a space to learn from people actually in the thick of it.

Product Ops HQ is brought to you by Dragonboat, the product operating system that helps teams and AI agents decide, plan, and act across the full PDLC—aligning strategy, execution, and outcomes at AI speed and scale.

Key Takeaways

How Sayari Conducts Agentic PDLC/SDLC

Neha shared Sayari transitioned from standard scrum and scaled agile methodologies toward an Agentic SDLC. This workflow integrates Claude, Dragonboat, and Jira via Claude’s Model Context Protocol (MCP) to handle intake, strategic breakdown, and execution tracking in a highly automated, yet human-governed loop.

The step-by-step workflow functions as follows:

Step 1: Unstructured Document & Annotation Ingestion

Rather than manually drafting detailed requirements and roadmaps from scratch, product managers feed raw, unstructured team discussions or documents directly to Claude.

“If we are annotating anything as part of Google Meets or if we have any other PRD documents, we were just feeding that into Claude and we were asking Claude to break that down into initiatives, features, and story.”

Step 2: Automated Strategic Breakdown & Cross-Tool Creation

Using Claude’s Model Context Protocol (MCP), Claude creates the strategic hierarchy inside Dragonboat and links it seamlessly down to Jira.

“The initiatives and features were getting created in Dragonboat and they were also being pushed from Dragonboat to Jira so the integration key was also getting established.”

“They are creating the items in Dragonboat via MCP, then they are breaking those items down, updating the statuses of Dragonboat work items in Claude, and then they’re pushing the work items in Jira.”

Step 3: Reconciling Existing Execution Back to Strategy

When developers spin up execution-level items directly in Jira without prior strategic alignment, Neha’s team uses Claude to retroactively tie them to the strategy layer.

“…And then if the work item already exists in Jira, via MCP, then they’re asking Claude to connect it back to Dragonboat, so it’s doing all of that pretty seamlessly.”

Step 4: Mandatory Human Governance & Data Validation

Despite the extreme automation and efficiency, Neha establishes human oversight as an uncompromisable gate before execution.

“Human governance doesn’t go away, right? You definitely ensure that you are utilizing AI for efficiency, but what I am ensuring with my product team, with my engineering teams, even with my go-to-market teams, is: guys, okay, now you have created this agentic feature and story, please review and please ensure that you are okay with this. It shouldn’t just directly go to developers for the execution without having the human governance review and finalization.”

“Every time, I at least find one or two discrepancies, so it is very important to do the human governance on that.”

Step 5: Shift to Kanban-Style Continuous Execution

Because the creation of stories and features has been highly compressed, the traditional, rigid sprint framework became a bottleneck. Sayari has shifted execution to support a more fluid flow.

“…If somebody asks me how processes are changing in AI world, I am saying yes, there is a thing like Agentic SDLC now, right? We have to go back to doing the more Kanban fashion of execution where the cycle time and WIP [work-in-progress] is becoming the real thing rather than before.”

The Role of Dragonboat in Agentic PDLC: Before and After Jira

Neha directly and indirectly addresses the misconception that Dragonboat is just a passive “passthrough” to Jira. Instead, she highlights Dragonboat as a critical, multi-directional strategic anchor.

Before Jira: Commercial Intake, OKRs, and Initiative Mapping

Jira is designed for team-level engineering execution, meaning it is blind to strategic ideation and commercial customer inputs. Dragonboat acts as the strategic foundation before execution begins.

• Centralizing Product Strategy and Intake: Neha’s product leadership specifically needed to manage early-stage items and corporate goals alongside developer work.
  

  “My product leadership wanted to manage the intake also in the same tool where we would be managing the strategic level, the OKR level items, and then the higher level initiative level work items which eventually gets broken into features and stories in Jira.”

• Integrating Salesforce Commercial Intake: Dragonboat uniquely bridged the gap between commercial sales requests and active development.
  

  “I was looking for a tool which is not only connecting the intake from Salesforce but also the execution happening like a team-level execution happening in Jira, and then bringing everything together as part of one tool. And I can confidently say that after doing that research, Dragonboat was the only tool which provided me that.”

After Jira: Strategic Reconciliation and Agentic Reporting

Once engineers execute in Jira, Dragonboat acts as the portfolio’s core database to synthesize execution outcomes and report back to the enterprise.

• Dynamic Portfolio Reconciliation: Using Claude’s MCP connection, execution data from Jira is constantly mapped back to Dragonboat’s high-level objectives.
• Automated Executive Status Reporting: Instead of PMs manually compiling slides or searching Jira boards, Neha uses Claude to read Dragonboat’s strategic fields and produce cadence reports directly for executives.
  

  “Now what I’ve started doing is really doing agentic way of reporting—meaning all the items, I’m ensuring there are certain fields which are exposed to MCP are being read in my reports and the scheduled-based cadence reports are being produced.”

• An Unchanged Enterprise Need: Neha notes that while AI speeds up execution, the foundational organizational requirements for alignment do not vanish.
  

  “The need of alignment still stays as is, the need of transparency still stays as is for any other enterprise, how it used to be before.”

How Faster Execution Drives C-Level Strategic Pressure and Ground Implications

The adoption of an Agentic SDLC has dramatically compressed development times, shifting the organizational bottleneck from “how fast can we build” to “how fast can we align”.

The Compression: Months to Weeks

By leveraging AI tools, Neha’s teams can now validate a feature’s direction in a fraction of the time.

“Before, on any bigger features, if we would have implemented that feature in a span of a month, we are able to get to the decent direction of where that feature execution is heading within a span of a week.”

The Resulting Pressure on C-Level Strategy

When execution happens at rapid speeds, C-level executives can no longer afford to review and pivot strategy on a slow, quarterly basis. They are placed under massive pressure to constantly re-strategize.

“…What that’s doing is creating more pressure and more need of constantly strategizing the next steps of the enterprise at the C-level execs, because now they are getting the execution outcomes a little bit faster than before, and whether or not they are ready to head in the same direction versus a different direction—kind of quickly pivoting…”

The New Implications for Teams on the Ground

This high-speed feedback loop creates a challenging psychological and operational shift for the execution teams.

• Strategic Fatigue: Even though execution teams are the ones enabling these massive efficiency gains, they are not accustomed to continuous strategic redirection.
  

  “…That is creating a different kind of mindset with the execution team because, even though they are the ones creating that efficiency, they are not that used to changing the strategy quite often.”

• The Shift in Operational Metrics: Rather than traditional velocity or milestone tracking, Neha focuses on measuring the rate of pivots and tracking execution flow using Kanban cycle time and WIP.
  

  “Measuring the pivots, measuring the execution to the strategy, and how quickly the strategy is pivoting is what we are seeing at the ground level, rather than relying on any traditional specific metrics like before… Cycle time and WIP is becoming the real thing.”