Junior AI & Automation Specialist

Location: London (Chancery Lane)

Working pattern: Full time, hybrid (circa 2/3 days in office)

 

About the role

Accordience Group is the parent company of four specialist communications agencies - Grayling, CDR, Red Consultancy, and Cirkle - operating across 30+ offices worldwide.

We are building custom AI tools across the group. We have strong demand and clear ideas; we need someone to turn them into working products that our agencies and clients can use.

This is a hands-on builder's role. You'll spend most of your time shipping - agents, internal tools, automations, reporting workflows, and the occasional client-facing MVP. You'll scope work with stakeholders, then own delivery end-to-end. The role may also involve working directly with clients to help embed AI within their communications functions.

 

What you'll do

• Build and ship custom internal AI tools, agents, and automations that the agencies use day to day
• Automate reporting, monitoring, and workflow tasks across the group
• Prototype and build MVPs of client-facing AI products, working alongside the AI team
• Work with internal teams and, where appropriate, clients to embed AI within communications workflows and functions
• Scope use cases with stakeholders across the agencies and translate them into working solutions
• Work directly with LLM APIs (OpenAI, Anthropic) and managed AI services on AWS and Azure
• Deploy serverless functions, wire up APIs, data sources, auth, and light front-ends around what you build
• Use workflow automation platforms (n8n, Make) for the jobs that don't need code
• Maintain and improve the tools you ship - and the ones already in flight
• Work with Group IT to stay inside approved platforms and security standards

 

What we're looking for

You've shaped real things, you can show us what they do, and you can explain why you made the choices you made.

We’re also open to recent graduates with a degree in Computer Science, Data Science, Software Engineering or a related field, ideally with some relevant work experience or internship experience.

 

You'll have:

• Hands-on experience building agents, workflows, or automations - ideally inside a company rather than just side projects
• Comfort with Python and scripting, and the ability to wire things together via REST APIs, webhooks, auth, and permissions
• Direct experience with LLM APIs (OpenAI, Anthropic, or similar) - prompts, tool use, agent design
• Hands-on with at least one cloud (AWS or Azure): deploying serverless functions (Lambda / Azure Functions), using managed AI services (Bedrock, Azure OpenAI), basic storage and auth
• Working knowledge of workflow automation platforms (n8n, Make, or equivalent)
• Enough front-end (HTML, React, CSS/Tailwind, or similar) to build a usable interface around what you've built
• Working knowledge of SQL and relational databases (e.g. PostgreSQL, T-SQL / SQL Server, or similar), including enough to query, structure, and connect data sources
• Experience with authentication and access management patterns (e.g. OAuth/OIDC; Auth0, Microsoft Entra ID, or equivalent)
• Git/GitHub and good repo hygiene
• A pragmatic, ship-it mindset - you'd rather get a v1 in front of users than perfect a v0 in private
• Strong communication skills, especially with non-technical teams

Useful but not essential:

• Multi-cloud exposure (AWS and Azure)
• Retrieval / RAG patterns - vector stores (Pinecone, pgvector), Azure AI Search, or similar
• Airtable or similar as a structured data layer
• Background in marketing services, PR, or communications
• Experience working with a Group IT or governance function
• Experience working directly with clients or embedded teams in a forward-deployed / solutions engineering capacity

 

How to apply

To apply, send us:

• Your CV
• A few lines on similar work you've done and what you took from it
• A link or demo of something you have built OR a short take on how you would build this: weekly client coverage, pulled and summarised for sentiment and themes, into a shared doc every Monday morning