# The hire I keep recommending: a practitioner ops lead, not an AI engineer Source: https://getstartedwithai.net/blog/the-hire-i-keep-recommending Site: Get Started With AI Tagline: Build AI automations that actually run your business. Audience: solo professionals, freelancers, consultants and AI implementers in training Contact: hello@getstartedwithai.co.uk This Markdown companion is generated from the prerendered page for AI crawlers. # The hire I keep recommending: a practitioner ops lead, not an AI engineer Do I need an AI engineer or an AI consultant: for most UK SMBs in 2026, neither. The right hire is a practitioner ops lead, an internal role at £45,000 to £75,000 a year that runs the WIRE Method on weekly workflows. AI engineers cost £80,000 to £130,000 and build models the SMB does not need. AI consultants cost £600 to £2,500 a day and leave at the end of the engagement. The practitioner ops lead is the role that owns the AI builds after the consultant has gone home. This post is the difference between the three roles, the cost comparison in pounds, the job spec template, and the rare cases where an AI engineer is actually the right answer. ## What an AI engineer does - Builds and maintains custom AI models for the business. - Manages infrastructure (training pipelines, inference, vector stores). - Writes Python and SQL daily, often in a cloud environment. - Designs prompt orchestration and multi-step agent flows. - Costs £90,000 to £130,000 a year in the UK in 2026, plus equipment and tooling. ## What a practitioner ops lead does - Runs the WIRE Method audit on weekly workflows that the team already does manually. - Configures Custom GPTs, Zaps and Make scenarios, plus Notion or Airtable setups. - Writes handover documents, runs the fortnight of output review, manages the named owner cadence. - Trains the rest of the team on the deployed builds and produces written escalation paths. - Costs £45,000 to £65,000 a year in the UK in 2026, often working from existing operations or office manager titles. ## Cost comparison side by side An AI engineer at £110,000 fully loaded costs the business about £140,000 a year once employer costs and equipment are included. A practitioner ops lead at £55,000 fully loaded costs about £70,000 a year. The annual difference of £70,000 buys roughly two implementation sprints from a working AI consultant or about 15 audit engagements. For most SMB AI work the engineer is over-specified and the audit-plus-consultant route plus the ops lead inside the team produces more shipped builds for less spend. The exceptions are real but narrow; most SMB founders asking the question can save themselves 18 months of recruiting time by hiring the ops lead first. ## Why most small businesses need the second role Most small business AI work is connecting existing tools to existing workflows. That work is operations work. An engineer hired for operations work tends to over-engineer: build infrastructure the business does not need, write code where a Zap or Make scenario would do, and ship deliverables that do not match the operational shape of the business. The ops lead is the right shape because the work is the right shape. The deeper version of the same point: SMB AI value comes from owning the builds after they ship, which is operations work, not from training new models, which is engineering work. Match the hire to the value. ## When you do need an AI engineer Three patterns where an AI engineer is the right hire for a small business. The first is a regulated environment where the model needs to run on private infrastructure for compliance reasons (financial services, healthcare, defence supply chain). The second is a product business where the AI itself is the product (an AI feature inside a SaaS product, an AI tool sold to other businesses). The third is a research-stage business where novel models are part of what the business sells to customers. Outside these three patterns, the AI engineer hire is almost always early. The honest practitioner read: in 120+ implementations, fewer than five clients needed an in-house AI engineer in the first 18 months. ## Job spec template for a practitioner ops lead - Title: AI Operations Lead (or Operations Lead with AI remit). - Reports to: COO, Head of Operations, or Founder if neither role exists. - Headline responsibility: own the AI workflows in the business; run the audit, identify, run, embed cycle on at least four workflows a year. - Required skills: process documentation, prompt writing in plain English, Zapier or Make at the configuration level, Notion or Airtable as a knowledge base, written communication. - Useful skills: SQL at a query level, Python at a script-reading level, basic understanding of LLMs but not training experience. - Excluded skills: model training, infrastructure design, vector store management. If the role needs these skills, hire an engineer instead and rescope. - Salary band UK 2026: £45,000 to £65,000 plus the standard SMB benefits package. - First six month deliverables: three shipped builds, written handover documents on each, named owner cadence in operation, fortnight of output review on every build. ## How to hire well: a paid trial beats a CV - Interview for shipped builds, not certifications. The portfolio matters more than the qualification. - Run a paid one-week WIRE Method audit on a real workflow as the final stage of the interview process. - Score the trial on three criteria: process clarity (can the candidate document the workflow in one page), owner identification (can the candidate name who would run the build), build feasibility (can the candidate sketch the smallest version that would ship in a week). - Do not interview for AI knowledge in the abstract. Test on the work. - Pay for the trial. A free trial attracts the wrong candidates and signals the role is not respected. ## The headcount-of-one play For a UK SMB under 30 staff, a single practitioner ops lead can run an AI implementation programme that produces four to six builds a year, owns the existing builds, and trains the rest of the team. The right structure is a single hire at £55,000 plus a fractional AI consultant on a small monthly retainer (£1,500 to £4,000 a month) for the bigger audits and the harder builds. Total annual spend: about £100,000 fully loaded. Output: a working AI implementation programme covering most of the operational workflows in the business. The structure I recommend most often is exactly this one, and it produces the highest ratio of value to spend in SMB AI work. ## Build your first AI system live with Richard A 60-minute live build with Richard Batt. Bring one repeatable task. Leave with a mapped automation, success metrics and a deploy plan you can test the next morning. Free seat, no upsell. ### Live build session The next live session covers the WIRE Method end to end on a real audience build. [Join the next session](/webinar) ## Questions people ask before they start ### undefined undefined ### undefined undefined ### undefined undefined ### undefined undefined ### undefined undefined ### undefined undefined