Rise of the Knowledge Engineer

Agents are one of the largest audiences for company knowledge. In July 2026, they accounted for 66% of measured web traffic across documentation powered by Mintlify.

There are people behind these agents. They've delegated their reading and, increasingly, their actions to AI tools. Developers ask coding agents to integrate APIs. Buyers use chat agents to compare products and even purchase them. Customers rely on support agents to solve problems.

For anyone responsible for documentation, help content, or internal knowledge, the job now extends beyond publishing accurate pages. Companies need systems that keep knowledge accurate, accessible, and consistent wherever people and agents work. The emerging role responsible for those systems is the knowledge engineer.

The cost of knowledge agents shouldn't trust

In 2024, an Air Canada chatbot incorrectly told a customer they could claim a discounted fare after completing their travel. When the airline refused the refund, the customer brought the dispute to British Columbia's Civil Resolution Tribunal.

The tribunal found Air Canada liable, concluding that the airline was responsible for information provided across its website, including by its chatbot.

Most agent failures aren't litigated. An agent can't find an answer about one product, so it recommends a competitor. It retrieves an obsolete API parameter and generates broken code. It finds conflicting policies and chooses the wrong one. Once an agent accepts incorrect information as authoritative, it can codify the mistake in generated code, support answers, and automated workflows, reproducing one error across thousands of downstream interactions.

Knowledge quality is an operational concern.

Knowledge work is moving from pages to systems

Technical writers, knowledge base managers, product educators, and support teams have traditionally focused on producing and maintaining content. That work remains important, but the scope is expanding.

David Hou, Product Education Lead at Decagon, describes the shift:

Agents are the new audience for your knowledge, and that comes with real product requirements: for how it's shaped, what it contains, and how retrievable it is, so the agent can perform meaningful tasks end-to-end for users. Building the knowledge layer takes more than technical infrastructure. It takes people and operating infrastructure too — the whole organization has to buy in, or a real source of truth never happens.

Fabrizio Ferri Benedetti, Content Engineer at Elastic and author at passo.uno, describes the emerging role as a "context curator": "a technical writer who is able to orchestrate and execute a content strategy around both human and AI needs or even focused on AI alone".

He says, "It doesn't matter if most of your documentation is generated by AI if you can ensure that it's done well and that it will produce the intended effect in LLMs. That content is the LEGO folks will use to build code cathedrals with AI's help. Once you've ensured that the pieces are accurate and easy for LLMs to access, that mission is complete. The other mission is helping humans, who need an opinionated view of the pieces and their possibilities"

What a knowledge engineer does

Knowledge engineering keeps the information a company uses internally and shares externally consistent across people, systems, and AI agents.

"Successful knowledge engineers connect the different sources of information spread across an organization and turn it into something useful," says Ethan Palm, Knowledge Engineer at Mintlify. "Docs and public content are key for your users, but internally it's all the knowledge scattered across different teams and tools. You need to build systems where information is trusted and accurate for your users and your team."

Knowledge engineers design the infrastructure and processes behind content. Their responsibilities include:

• Establishing authoritative sources and clear owners.
• Connecting product releases, code changes, support tickets, and feedback to content updates.
• Publishing knowledge as structured pages, Markdown, llms.txt indexes, APIs, and MCP tools.
• Defining terminology, review requirements, permissions, and freshness standards.
• Measuring what agents search for, which questions go unanswered, and where retrieval fails.

Tools can automate parts of this loop. Mintlify can serve Markdown to agents, generate llms.txt indexes, and distinguish human and agent traffic. The knowledge engineer determines what is authoritative, how it stays accurate, and when a person must review it.

Agents change what humans need

People remain an important audience, but they need something different from documentation.

Agents benefit from exhaustive, structured information such as parameters, constraints, schemas, and procedures. Humans need help understanding why a system works, which approach to choose, and what trade-offs matter.

Drew Breunig writes that "Humans don't require exhaustive documentation, they require mental models. Agents can find the details." He recommends that documentation should "help people ask better questions and write better instructions."

Knowledge engineers give agents dependable source material while helping people understand and apply it.

The benefits of a knowledge engineering mindset

Keep knowledge current and accessible

Updates can flow from their source instead of depending on someone remembering every downstream page. Automated detection and drafting also reduce the delay between a product change and its explanation. Coinbase reduced documentation update time from more than 20 minutes to under 60 seconds.

Structured content, clear navigation, Markdown, llms.txt, and MCP also help agents retrieve the right information without guessing.

Turn gaps into improvements

Agent searches and unanswered questions reveal gaps that page-view analytics miss. Teams can see where agents struggle, prioritize the missing information, and measure whether an update solved the problem.

Enable teams to prioritize

Subject-matter experts can review proposed changes while automated systems handle repetitive drafting and distribution. Writers and engineers gain more time for customer experience, adoption, and product quality. HubSpot cut the engineering resources devoted to documentation infrastructure by 50%, freeing the team to focus on developer growth and retention.

Knowledge needs an owner

Knowledge engineering expands the scope of technical writing by connecting content creation and governance directly to the systems that engineering, design, and product teams use.

Companies can start by assigning accountability before creating a new title. Someone must identify authoritative sources, connect updates to their origins, make knowledge retrievable, and measure where the system fails.

"There are so many skills that go into being a strong knowledge engineer," says Palm. "Don't limit yourself to one particular title or list of tasks. You need to understand the problems for how you and your users need to access your knowledge. Then find the people who can steward it."

The advantage will go to companies that turn change into trusted, usable knowledge fastest for both people and the agents acting on their behalf.