AI Knowledge Operations Manager (m/w/d)

Role Overview

You enjoy analyzing complex support conversations to identify the root causes of issues. You think systematically and process-oriented, preferring to build scalable solutions rather than repeatedly answering the same questions. You have a genuine eagerness to work directly with AI agents, not just discuss them.

At Candis, an AI agent integrated into Intercom currently handles a significant portion of customer support. Although the agent performs many tasks, there is currently no single person accountable for its effectiveness — this is where you come in.

Why This Position Exists

The AI agent’s quality depends entirely on the underlying knowledge base. Presently, this knowledge is incomplete, outdated in places, and updates to product features outpace content revisions. As a result, the agent participates in many conversations but cannot resolve all issues, diminishing customer experience. Currently, the agent manages approximately 64% of support conversations and successfully resolves about 69% of them. We aim to improve these metrics.

Maintaining content alongside ongoing product development proved ineffective without dedicated responsibility. You will take ownership of maintaining and updating knowledge, define when the agent should respond or escalate issues, and establish processes to keep content aligned with the product. Your work will also support internal teams that depend on this knowledge daily.

Why This Job Is Exciting

This position has a significant impact: you shape how the majority of customers experience our support. Our support service is already high-quality, but there is room for improvement in the AI agent, which affects our customer experience scores that have recently declined slightly. Your job is to enhance these results using modern tools and technologies. Sales and product managers also rely on your work internally. The role offers substantial freedom to experiment and learn how to set up, supervise, and control AI agents. The Intercom platform is widely appreciated, and you will enjoy working with it.

Key Challenges

• Building and continuously improving product knowledge is a large undertaking while simultaneously learning on the job.
• Prioritizing issues effectively is critical due to high content volume; you cannot fix everything at once.
• Bringing structure and constantly choosing the most impactful next topic.
• Cooperating across functions, including product teams, engineering, and increasingly sales, especially regarding chatbot deployment.
• Investigating parts of documentation that are outdated or incomplete and where original contributors may no longer be present, requiring detective-like efforts to understand functionality.

First Year Goals

• Within 30 days: immerse yourself in support activities, shadow sales demos and onboarding, study conversations and the help center, and independently handle a few support cases. Know where to find answers for each topic and select a manageable topic for a focused deep dive.
• Within 60 days: master Intercom as your main tool and establish a basic categorization to quantify conversation volume.
• Within 90 days: develop strong product understanding and detailed knowledge on your first assigned topic. Identify causes of problems, responsible parties, and process gaps. Start tuning content and the AI agent, measuring KPI improvements while conversation volume increases.
• Within one year: fully renovate the help center to cover approximately 90% of relevant cases, ensure dependable escalation procedures, and implement AI-driven processes that keep content current despite product changes.

Ideal Qualifications

• Strong analytical skills to decipher disorganized conversations and detect true issues beyond surface-level ratings.
• Systemic and process-focused thinking to design sustainable Intercom workflows that remain effective despite product updates.
• Empathy toward customers to recognize when interactions leave users feeling unheard, then improve processes to prevent recurrence.
• Experience with AI-first workflows by creating semi-automated, AI-supported content and support processes, reviewing and refining AI-generated suggestions rather than writing manually.
• Fluency in German at a C2 level, with good English communication skills. Approximately 90% of support conversations occur in German.
• Residence in Berlin to work hybrid or on-site.

Additional Advantages

• Hands-on experience with Intercom, including implementing or maintaining AI agents.
• Background in knowledge management, content management, or technical writing.
• Strong writing skills to effectively edit and improve AI-generated articles.
• Expertise organizing large content libraries for both large language models and human users.
• Experience in support, service design, or support operations.
• Close collaboration with product and engineering teams.

About Candis

Candis is an emerging fintech company focused on transforming accounting and bookkeeping practices. Our platform helps businesses regain control over their invoice workflows—including invoice collection, automated data extraction, account assignment, and payment processing. With a modern, intuitive app design, we streamline workflows not only for accountants but across teams.

We empower finance teams with software that generates insightful financial analytics, aids decision-making, supports company growth, and promotes sustainability by minimizing paper usage in accounting.