How I approach product management with the most recent launch

Hi, I am Berk, product manager with 10 years of experience building AI native products, AI agents, advertisement technologies, B2B SaaS, procurement and supply chain technologies, and mobile apps.

I know it's not easy to refer someone you never work with, so instead of just sending my CV with a cultural outage from LinkedIn, I want to record a video to show how I turn an ambiguous problem to an end product that our customers love.

Hopefully by end of this video you will have a better sense of the kind of PM I am, the value I could bring your company and feel confident referring me if you see a fit.

So here's a recent example.

Paktum had strong enterprise demand, but the product was quite difficult to scale.

Through customer conversations, deployment retrospectives, sales feedback and negotiation data.

I found three connected problems.

Sales struggled to clearly position the right use cases.

Customer specific configuration made deployment slow with an average time to life of four and a half months.

Our deterministic flows couldn't adapt sufficiently to the constraints and the preferences suppliers revealed during a conversation.

My mandate was turn these bespoke deployments into repeatable products.

I reframed the products around the complete supplier engagement journey, discovering commercial opportunities, selecting right suppliers, configuring customers policies, and executing autonomous negotiations.

I translated these findings into product model that engineering, commercial and deployment teams could reason about.

Together we define reusable capabilities, configuration boundaries, and a shared language for explaining the platform to our customers.

I built this prototype to make the interaction model tangible and explore an important product.

How can an agent adapt its strategy while remaining predictable, understandable, and bounded by customer policy?

Here the agent first confirms that the supplier's current payment terms is 30 days.

It then opens with a proposal of 90 days with a 3% discount.

The supplier rejects any extension to the payment terms, but indicates that they are open to discussing a general discount.

The important product decision here how the agent interprets the supplier constraints in this case payment terms and uses them to adopt a negotiation strategy.

We can see the transition here.

The payment terms now lock to the 30 days, but while the discount is repriced to 4 and a half percent to preserves as much as possible to the original commercial value.

When the supplier counter offers with 2%, the agent responds with a final offer with 3.5% without changing the payment term.

Now the payment term is locked and the supplier is agreed to reach an agreement with the 3.5% discount contribution was to identify why the deterministic flow was limiting.

Conversion model how supplier signals should update the negotiation state and define which decisions could be agentic and which need to be remain deterministic.

I translated that logic into experience where agent's adaptation was visible and understandable.

As you see, for the first iteration I deliberately focused on a small number of high value negotiation signals representation rather than attempting to handle every possible supplier response.

I worked closely with engineering to turn these reusable states and routing rules, then tested the experience with customers and on our commercial terms.

This wasn't only a strategy or a exercise.

The resulting capabilities changed how we configured, positioned and executed supplier engagements across broader platform work.

We increased the number of supplier and engagement use cases per customer by 50%, reduce the time to life from four and a half months to two months, and increase the share of initiated negotiations reaching agreements by 25%.

That's generally how I like to work as a product manager.

I enjoy continuous discovery talking customers, turning massive problems into clear product bets, then working closely with the engineering and design team to ship, learn and iterate.

I am also very hands on, especially with AI.

I like building prototyping fast, experimenting with new tools and models and using working software to learn faster rather than relying on spec or slides.

I am also a former founder myself, so I tend to think beyond just building product.

I care about whether we are solving the right problem, how it creates the value for our business, and ultimately whether the customers are willing to use and pay for it.

Hopefully this gives you a better sense of how I work and what I could bring to the role that you are hiring for.

If you think my experience and the way I work could be a good match, I would really appreciate your referral or simply a chance to have a conversation with the team.

Thank you for watching and hope to talk to you soon.