Owning Product at Raffle AI

I own product direction at Raffle AI, a small B2B AI SaaS company. Since my role was formalised in April 2025, I have set roadmap priorities, scope, sequencing, and release decisions while working closely with engineering, commercial teams, leadership, and customers.

The company context

Raffle AI builds search and chat products for enterprises. A small cross-functional team serves customers while developing the shared platform, so product decisions have to balance customer needs, commercial priorities, technical constraints, and the capacity required to improve the underlying product.

When I began coordinating product work in January 2025, no one held a dedicated Product Manager role. My authority was formalised in April: I became responsible for roadmap priorities, scope, sequencing, and release approval. The CTO retained formal resource allocation, and the CEO agreed quarterly commitments and could overrule product decisions.

My scope

My role connects the people closest to different parts of the product: customers and customer success surface needs; Sales and partnerships bring commercial context; engineering and data colleagues shape technical options; leadership sets company priorities. I turn those inputs into product direction and make the scope and sequencing decisions needed to move work into delivery.

That scope has included four complementary areas of work:

• Making a rebuild worth funding. I combined an architectural rebuild of the customer-installed widgets with the redesign and backlog features the business was asking for, so one approval paid for both, and set the requirements the new architecture had to meet.
• Rebuilding the chat around what people were asking it to do. A year of real conversations showed that many of the people using the chat were not searching — they were trying to do things. I proposed inverting the architecture so the chat became an agent with tools and search became one of them, ran the discovery, and built the proof of concept; the engineering and data science teams built the production version.
• Scaling a partner channel. I defined and led the product work for infrastructure that let external partners onboard and manage customer accounts without relying on repeated one-off configuration.
• Making product commitments dependable. I introduced a shared roadmap, clearer delivery conventions, and a longer planning horizon to replace week-to-week prioritisation with explicit quarterly commitments.

The focused case studies cover the decisions, trade-offs, execution, and limits behind each area without repeating the company context in full.

Two outcome signals

Infrastructure for 200+ partner-managed accounts

The partner platform’s core workflows shipped in January 2026, followed by a shared monitoring dashboard in Q2. The platform enabled external partners to onboard and manage more than 200 paying customer accounts through reusable widget templates, configurable commercial plans, and shared account visibility.

Sales and the partners drove customer acquisition. My contribution was defining and leading the product work that made the channel operable at that scale; engineering built and shipped the platform.

More realistic quarterly commitments

The team completed 4 of 14 roadmap projects in their promised quarter in Q3 2025 and 9 of 10 in Q2 2026. Progress between those points was not linear: the full sequence was 4 of 14, 4 of 6, 5 of 8, and 9 of 10.

“Delivered in the promised quarter” is the definition used here. No project was removed from the denominator, and late projects moved into the following quarter. I owned prioritisation, planning, and sequencing; the CEO agreed the commitments and engineering delivered the work.

How to read the results

My responsibility was to make product direction explicit: choose what entered the roadmap, define and negotiate scope, sequence the work, and approve releases. Engineering owned implementation, and Sales, partners, leadership, customers, and customer-facing colleagues supplied essential commercial, technical, and user input alongside it — these outcomes reflect that cross-functional effort, not a Product Manager working alone.

This overview establishes that company-level scope. The focused stories carry the detailed evidence: what the team knew, which options we considered, what I decided, what each choice traded away, and where the available evidence still has limits.