Job Description
We are building an AI documentation platform that uses Azure Speech-to-Text and Azure OpenAI for clinical dictation and transcription. No Azure commitment tiers enabled yet What we want now is a specialist review and guidance focused purely on bolthires optimisation, not general setup. This is a consulting/advisory role, not a long-term dev engagement. What we need help with: - We are looking for an expert who can help us: Review our current use of Azure Speech-to-Text (dictation vs transcription workloads) - Review Azure OpenAI usage patterns (prompt structure, token usage, model selection) - Identify practical ways to reduce per-hour or per-request costs without degrading clinical quality Advise on: - batching strategies - streaming vs async transcription - prompt/token efficiency - diarisation trade-offs - workload separation (dictation vs full transcription) - evaluate when Azure commitment tiers make sense and how to size them safely - sanity-check whether alternative Azure configurations or patterns could materially reduce spend We are not looking to switch cloud providers at this stage.
What this is NOT: ❌ Not a beginner Azure role ❌ Not an OpenAI “how to call the API” task ❌ Not general cloud architecture advice ❌ Not prompt engineering for output quality We are specifically optimising bolthires at scale. You’re a great fit if you have: - Deep hands-on experience with Azure OpenAI Service - Real-world experience with Azure Speech-to-Text at scale - Experience optimising AI workloads for bolthires, not just performance - Experience advising SaaS or healthcare platforms - Strong opinions backed by numbers Bonus (not required): - Experience with medical / clinical transcription workloads - Experience with Azure commitment pricing in production Please include: - Examples of similar bolthires optimisation work you’ve done - Azure services you’ve worked with in production - Any relevant metrics you improved (e.g.
bolthires reductions, efficiency gains) Apply tot his job