Product Manager

At Arbisoft, we partner with companies at different stages of growth to build products that solve meaningful problems. As a Product Manager, you'll work closely with clients, engineers, designers and stakeholders to shape product direction, validate ideas, and deliver outcomes that matter succeeding not by how loudly you lead, but by how seamlessly you become a trusted extension of the teams you support.

WHAT YOU'LL DO

• Work with clients and cross-functional teams to understand business goals, users and market opportunities.
• Drive product discovery through research, analysis and validation.
• Turn business objectives into requirements and roadmaps precise enough that engineers, and their AI tools, can execute without guessing.
• Prototype ideas with AI tools before committing engineering time.
• Define acceptance criteria and evaluation for AI-powered features, whose behaviour is probabilistic rather than fixed.
• Prioritise with engineering and design; use data and experimentation to decide.
• Track performance, define success metrics, and communicate trade-offs to stakeholders.
• Adapt quickly across industries and business models as client needs evolve.
  
  
  

WHAT WE'RE LOOKING FOR

• Typically 4+ years in product management.
• Strong product discovery and strategy skills; comfortable with technical trade-offs.
• Hands-on with modern AI tools, with a working sense of what LLM features can and can't do reliably.
• Demonstrated use of data and experimentation to drive decisions.
• Strong business acumen; picks up unfamiliar domains fast.
• Excellent written and verbal communication — your writing is the spec.
  
  
  

NICE TO HAVE

• SaaS or technology product experience across multiple industries.
• Experience shipping and measuring an AI-powered feature.
  
  
  

WHAT SUCCESS LOOKS LIKE

• You quickly build context around new businesses and products.
• Clients trust your judgment and treat you as part of their team.
• Your decisions are grounded in customer insight, data and business impact.
• Engineers rarely need to come back and ask what you meant.
  
  
  

HOW WE ASSESS
Judgment over recall: a practical task with AI tools allowed, then interviews on the trade-offs behind your decisions.