Phaenna is the AI consulting and advisory practice of Sophia Harris. One consultant runs it, from the first call through the diagnostic, the build and the follow-up. Nobody scopes the work and hands it to someone you have not met.
Before AI, Sophia's job was persuading people inside organisations to change how they worked. That is still most of this job: AI projects stall because nobody owned the output, or because the people expected to use the thing were told about it last. The work is to settle that during the build rather than discover it afterwards.
AI is neither frightening nor limitless. It is a set of tools with a real edge, and most of the value is in knowing where that edge falls.
One version of the story buys licences and then quietly discourages anyone from using them, and the other buys a platform before anyone has looked at the process it is meant to fix. The unglamorous work in between is finding out which processes would genuinely be better with these tools in them, and which would not. The second half of that answer is worth as much as the first.
Operations leaders, function heads and general managers responsible for how work gets done, with a budget and a process that is slower, more manual or more error-prone than it should be. Not teams looking for a tool demonstration.
Alongside client work, Sophia does model evaluation, debugging and rubric design for leading and frontier AI labs. She does not sell that work to clients. She does it because it keeps the advice matched to how these models actually behave rather than how they are marketed.
Training and mentoring delivered through partner organisations, alongside direct client engagements.
Claude, ChatGPT and Microsoft Copilot, built into the Google Workspace, Microsoft 365 and CRM environments teams already work in.
Want to know whether Phaenna is the right fit for your team?
Book a diagnostic