SVP of Product says traditional PM practices are “exactly wrong” thanks to AI
Discover how AI helps product teams accelerate discovery and product delivery while proving that strategy, customer insight, and judgment remain essential to success.
Ibrahim Bashir
SVP of Product at Ontra
Team Focus: Ibrahim Bashir leads a diverse team at Ontra, focusing on AI-driven product management and operations.
AI Impact: AI is reshaping traditional product management assumptions, altering how teams approach change and user engagement.
Development Efficiency: Bashir leverages AI to streamline product lifecycle processes, enhancing communication and reducing time from ideation to launch.
Customer Collaboration: Effective integration of AI has led to challenges in customer engagement and managing expectations during product rollouts.
Learning Mindset: Product leaders should experiment with AI continuously, adapting to new tools and approaches in product development.
Ibrahim Bashir previously led product initiatives at Box, Twitter, Amplitude, and Amazon, and he's currently the SVP of Product at Ontra. He's the creator of the Run the Business newsletter.
We sat down with Ibrahim to get a sense of how AI is changing his organization — and product management in general. Here's what he shared.
Building teams, shipping products, and repeating outcomes
I'm Ibrahim Bashir. I'm currently SVP of Product at Ontra, where I lead roughly 60 people across product management, user experience, data science, and operations.
Want more from The CPO Club?
Sign up for a free membership to complete reading this article:
Have an account? Log In
Ontra is an AI-powered legal operating system for private markets: We automate contracts, obligations, entity management, and compliance workflows across the full-fund lifecycle. Our customers are primarily private equity and venture capital firms, investment banks, direct lenders, and the law firms and service providers that support them. All told, we work with more than 1,000 firms globally, including nine of the top ten private equity firms.
Our delivery model is probably the most distinctive thing about us, and it's especially relevant to this conversation. We ship both traditional subscription SaaS products and a managed services offering where expert legal resources use AI to execute work directly on behalf of customers.
This dual model allows us to see, in real time, where AI can handle the full workflow end-to-end and where a human still needs to be in the loop. It provides an incredible vantage point for understanding how AI-native products get built and adopted in high-stakes enterprise settings.
Our delivery model is probably the most distinctive thing about us, and it’s especially relevant to this conversation.
Before Ontra, I scaled the flagship product at Amplitude, grew new businesses at Box, built service infrastructure at Twitter, and shipped in the Kindle ecosystem at Amazon.
Across all of it, I've focused on three things: building teams, shipping products, and repeating outcomes. I also write Run the Business, a Substack for B2B product leaders, which I use as a canvas to clarify my thinking. My goal is to connect lessons learned in practice, ideas absorbed from reading, and frameworks borrowed through collaboration.
Why AI made old product assumptions wrong
What makes this AI moment different, for me, isn't the model capability curve. It's that AI is quietly eroding most of the assumptions I'd internalized as a B2B product leader over the last decade.
Operating principles I used to recite without thinking aren't exactly wrong, but AI is reshaping them in real time.
Take change management. It involved reducing the learning curve so users pushed through a new tool's friction. In an AI-native product, the tool often isn't hard to use; the resistance is more existential — "Am I training my replacement?" Different root cause, different response.
So the real work of product leadership right now is less about adding AI to the roadmap and more about honestly auditing which of your priors still earn their keep.
Where AI yields the most ROI for PMs
I leverage AI for all facets of the product development lifecycle, but large-scale synthesis yields the most ROI:
- Summarizing themes across customer calls
- Generating variants of roadmap decks
- Turning strategy talking points into different altitudes of communication
- And so on
Judgment and interpretation work that lacks much precedent or context still primarily requires humans. But AI is becoming a good sounding board even in these scenarios.
How AI can reconcile diverse product perspectives
Here’s another benefit: AI simplifies the process of unifying and reconciling product perspectives across a company, reducing the friction that typically surfaces as a product nears launch.
Now, teams can more easily build a shared understanding of the problem, persona, solution, and differentiation much earlier using written and visualized artifacts, preventing misalignment later in the development lifecycle.
How product leaders can prepare for customer meetings with AI
When I'm preparing for a customer meeting, I stitch together all our breadcrumbs about the account across different systems, including call transcripts, to get a 360-degree picture of what they care about, progress on product feedback, and a dossier on key stakeholders.
The workflow involves Claude alongside tools like Gainsight, Gong, and Snowflake.
How AI impacts product cycle time
AI has encouraged me to be more ambitious and also revisit old ideas that perhaps would have been harder to explore previously.
It has also made me more amenable to quickly trying and killing product ideas. In our industry, even though we talk about product bets, we rarely start something that doesn't ship.
AI allows us to conduct more rapid brainstorming and prototyping, decreasing cycle time from idea to release.
But there's a flip side to that. Product teams must consider and respond to more ideation. Keeping up with the volume of ideas from across the company makes every week like hackweek.
AI allows us to conduct more rapid brainstorming and prototyping, decreasing cycle time from idea to release.
How customers can bottleneck AI's benefits
Now that AI has decreased our cycle times, we've come across a new issue.
Partner availability as co-developers is still limited. We sell into a niche buyer base, so their time and ability to participate in Beta programs are constrained. They can only absorb a limited amount of change management.
So, while AI has helped with volume, we are still figuring out its rollout and enablement.
How AI creates unrealistic expectations
AI creates a false sense that productionizing is easy, so expectations for shipping velocity and adoption are out of whack.
Expect faster brainstorming, but at some point, GA is a different bar.
Why AI's cost-effectiveness needs to be reviewed
We didn't discuss the total cost to build.
Many view AI as an almost free productivity boost, but eventually, you must compare token spend to people spend, and the math may not always favor AI.
For example, using Claude to monitor your Slack and read messages might not be cost-effective soon.
Why product leaders must roll up their sleeves and experiment with AI
I encourage product leaders to maintain a learning mindset, stay curious, and encourage experimentation. Don't be too precious about best practices from the prior era of software development.
I've had good success rolling up my sleeves and sharing my own hacking, incentivizing teams with fun awards for creativity, and carving out dedicated time for tinkering instead of focusing 100% on roadmap delivery.
Also, it’s okay to dabble in multiple tools for the same scenario. Models and products are evolving rapidly — no need to make premature lock-in decisions.
Follow along
You can follow Ibrahim Bashir's work on LinkedIn. And check out his Substack, Run the Business.
More expert interviews to come on The CPO Club!
