Responsible AI Policy

1. Purpose & Principles

Minima Maxima is committed to the safe, ethical and transparent use of Artificial Intelligence (AI).
Our mission is to help recruitment agencies and people-based businesses adopt modern AI tools in ways that are compliant, commercially beneficial and aligned with human values.

This policy outlines how we design, deploy and govern AI-enabled solutions for our clients.

Our guiding principles are:

  • Human-Centred: AI must enhance human judgement, not replace it.

  • Transparent: Outputs, logic and limitations must be clear and explainable.

  • Secure & Compliant: Data used in AI systems must adhere to GDPR, privacy laws and client governance.

  • Fair & Ethical: AI solutions must avoid reinforcing bias.

  • Accountable: We take responsibility for AI implementation and ensure clients remain in control.

 

2. Scope

This policy applies to:

  • All AI-enabled services offered by Minima Maxima

  • All advisory, configuration, automation, analytics and optimisation engagements involving AI

  • Third-party AI tools integrated into client ecosystems (e.g., copilots, automation logic, LLM workflows, retrieval systems, semantic search)

  • All data accessed, used or processed as part of any AI deployment

 

3. Human Oversight & Decision-Making

AI tools deployed or configured by Minima Maxima must never autonomously make decisions that have direct human, financial or legal consequences.

Examples of acceptable AI support:

  • Drafting communications

  • Generating shortlists

  • Summarising candidate information

  • Suggesting insights

  • Flagging risks, anomalies or gaps

  • Enhancing internal search and knowledge retrieval

 

Examples of prohibited autonomous actions:

  • Rejecting or excluding candidates

  • Making hiring or placement recommendations without human review

  • Triggering compliance or regulatory actions

  • Making contractual or financial decisions on behalf of clients

All AI-generated output must be reviewed and approved by a human before being used in any high-impact workflow.

 

4. Data Privacy, Security & Compliance

4.1 GDPR Compliance

  • Personal data is processed lawfully, fairly and transparently.

  • Data minimisation is enforced — only relevant information is used.

  • Clients retain full ownership of their data.

  • Data ingestion and processing follow client-specific retention rules.

  • All AI usage is logged and traceable.

4.2 Data Security

  • Secure authentication and access controls are mandatory.

  • Client PII is never shared with consumer-grade or non-compliant AI tools.

  • Enterprise-grade, GDPR-compliant AI services are used wherever possible.

4.3 Confidentiality

  • Client data must remain confidential and cannot be used to train public AI models.

  • Private model routing or environment isolation is used when available.

 

5. Bias, Fairness & Ethical Use

AI may unintentionally reinforce bias if not properly managed.

Minima Maxima ensures that:

  • AI-generated recommendations are assessed for fairness

  • Data sources used in retrieval or training are reviewed for bias patterns

  • Automated steps do not unfairly favour or exclude groups

  • Clients receive guidance on ethical risk areas, especially in recruitment and people-data workflows

  • Identified issues trigger immediate review and corrective action

We promote ethical handling of candidate and client data in alignment with DE&I principles.

 

6. Transparency & Explainability

AI systems deployed by Minima Maxima must be:

  • Understandable to users

  • Able to explain the reasoning behind outputs

  • Clear about limitations (including hallucination risks)

  • Configured to log relevant actions, prompts and outputs

  • Documented with usage guidelines for client teams

Users must always be aware when AI is being used and how its outputs should be interpreted.

 

7. Appropriate Use & Risk Boundaries

Minima Maxima advises clients not to use AI for:

  • High-risk compliance or regulatory decisions

  • Inferring sensitive or protected characteristics

  • Automated candidate ranking without oversight

  • Automated rejection or adverse messaging

  • Mental health, legal, financial or safeguarding assessments

AI must always act as a decision-support tool, not an autonomous decision-maker.

 

8. Monitoring, Review & Continuous Improvement

We continually evaluate AI systems to ensure they remain safe, compliant and effective.

This includes:

  • Regular audits of AI workflows

  • Monitoring output quality and error rates

  • Identifying new risks as models evolve

  • Ensuring alignment with emerging AI regulations and industry standards

  • Updating client governance frameworks as tools mature

 

9. Client Responsibilities

Clients using AI-enabled solutions are responsible for:

  • Maintaining internal data governance and compliance

  • Training staff on correct use of AI tools

  • Ensuring AI outputs are reviewed by appropriate team members

  • Reporting issues, anomalies or risks to Minima Maxima

  • Following agreed usage guidelines

Minima Maxima provides guidance, templates and training as needed.

 

10. Accountability & Contact

Minima Maxima is accountable for the design and configuration of AI systems deployed under our guidance.

We encourage clients to raise concerns, request audits or seek clarification at any time.

For Responsible AI enquiries:
Email: tony@minimamaxima.io
Website: minimamaxima.io

Automating VMS Job Workflows for Higher-Margin Growth

OVERVIEW

A specialist recruitment agency supplying contract talent into large enterprise clients relied heavily on VMS (Vendor Management System) job flow.
Despite strong demand, the business struggled to convert VMS roles efficiently due to:

  • High job volume
  • Short turnaround expectations
  • Intense competition
  • Low-margin roles consuming disproportionate recruiter time

 

Minima Maxima partnered with the agency to automate their VMS job workflows end-to-end, increasing speed, accuracy and conversion — while freeing consultants to focus on higher-margin activity.

THE CHALLENGE

  1. High Job Volume, Low Signal

 

VMS feeds were generating hundreds of roles a week, many of which:

  • Were duplicated
  • Had incomplete details
  • Involved outdated rates or job loads
  • Contained irrelevant market segments

 

Consultants were spending excessive time manually triaging jobs.

 

  1. Slow Reaction to New VMS Drops

 

When a job appeared:

  • It took too long to get into Bullhorn
  • Consultants manually checked skill fit
  • Prioritisation was inconsistent
  • Response times were slower than competitors

 

Speed-to-first-submission was a major weakness.

 

  1. Low Margin Roles Consuming High Margin Time

 

Recruiters were:

  • Rebuilding job details manually
  • Re-sourcing candidates already in the database
  • Managing compliance admin for low-value assignments
  • Burning hours on admin-heavy VMS accounts

 

This crushed productivity and hit overall GP per recruiter.

 

  1. Poor Visibility of Job Coverage

 

Leaders couldn’t see:

  • Which VMS roles were being worked
  • Which were stalled
  • Where leakage was happening
  • What revenue was left on the table

 

This meant decision-making was reactive, not strategic.

OUR APPROACH

  1. Mapping the VMS-to-Bullhorn Workflow

 

We ran a full process audit across:

  • VMS job ingestion
  • Bullhorn job creation
  • Matching & shortlisting
  • Compliance & documentation
  • Submission workflows

 

Every friction point and manual step was documented.

 

  1. Automated VMS Job Intake

 

We implemented a streamlined job ingestion framework:

  • Auto-deduplication of incoming jobs
  • Auto-tagging by skill, market and region
  • Validation checks for missing fields
  • Standardised job naming conventions
  • Job prioritisation rules (speed, margin, exclusivity)

 

This transformed VMS feeds into organised, high-quality job records ready to work instantly.

 

  1. Automated Candidate Matching

 

We designed a matching engine powered by:

  • Skill taxonomy
  • Historical placements
  • Search prerequisites
  • Availability markers
  • Automation-led shortlists

 

Candidates matching the role were automatically:

  • surfaced to recruiters
  • added to a shortlist
  • sent a quick-availability workflow
  • flagged to the relevant desk lead

 

Turnaround times dropped dramatically.

 

  1. Automating Low-Margin Admin

 

We deployed automations to remove low-value manual work:

  • Submittal preparation
  • Compliance reminders
  • Rate confirmation workflows
  • Candidate update sequences
  • Interview coordination steps

 

Recruiters reallocated hours toward higher-value clients and roles.

 

  1. Real-Time VMS Coverage Dashboard

 

We built an analytics layer that showed:

  • VMS job volume by vendor
  • Stage progression
  • Submission speed
  • Coverage gaps
  • Win/loss ratios
  • High-margin vs low-margin segmentation
  • Recruiter performance by account

 

This enabled leaders to make smarter decisions about:

  • Allocating resource
  • Prioritising vendors
  • Pricing and rate negotiations
  • Strategic focus areas

THE IMPACT

  1. Speed-to-Submission Improved by 40–60%

 

Automated matching and workflow triggers meant:

  • Job review → shortlist → submission happened in minutes
  • The agency beat competitors to first submission
  • Fill rates increased on exclusive and semi-exclusive roles

 

  1. Reduced Recruiter Admin Time

 

Recruiters gained 6–8 hours per week thanks to automated tasks.
This time was reinvested into:

  • High-margin clients
  • Deeper candidate engagement
  • Market mapping
  • BD activity

 

  1. Higher Fill Rates on Repeat VMS Roles

 

Automation ensured:

  • Immediate reuse of recently active candidates
  • Consistency in post-placement care
  • Stronger candidate relationships
  • Better compliance workflow adherence

 

This directly improved contractor GP.

 

  1. Low-Margin Roles Became Worth Working

 

Because the admin overhead was reduced, VMS roles:

  • Became economical to fulfil
  • No longer drained recruiter time
  • Contributed healthy incremental GP

 

The agency shifted the economics entirely.

 

  1. Data Became Actionable

 

With structured, clean, automation-ready data:

  • Leaders gained real visibility
  • Recruiters trusted Bullhorn again
  • Reporting aligned with operational reality

 

The business became measurably more predictable.

 

What This Enabled

 

By automating the entire VMS-to-placement lifecycle, the agency unlocked:

  • Faster role qualification
  • Higher productivity per recruiter
  • Consistent delivery on time-sensitive roles
  • A lower cost-to-serve
  • Higher overall GP
  • A fully scalable VMS delivery engine

 

The company now treats VMS workflows as a growth channel, not an operational burden.

CLIENT QUOTE

“Before the automation project, VMS roles were a drain. Now they’re one of our most profitable channels — our team is faster, more consistent, and actually enjoys working them.”

Turning a Global Tech Recruiter’s Data Into Deal Flow

OVERVIEW

A global technology recruitment company operating across the UK, EMEA and the US approached Minima Maxima with a familiar challenge:
tons of data, thousands of candidates, but no meaningful deal flow being generated from it.

Despite having years of candidate CVs, notes, outreach history and job records inside Bullhorn, the business struggled to leverage its data for sourcing, re-engagement or market insights.

Minima Maxima partnered with the leadership team to convert their legacy data into a high-performing revenue engine, powered by automation, analytics and an AI-ready data foundation.

THE CHALLENGE

The agency had grown fast — but without a consistent data strategy. As a result:

 

  1. Massive but messy data sets

 

  • Hundreds of thousands of candidate records
  • Duplicates across regions and business units
  • Quality varying significantly by market
  • Missing fields blocking sourcing and search

 

  1. Recruiters relying on LinkedIn instead of Bullhorn

 

  • Bullhorn wasn’t perceived as a reliable search tool
  • Searching the CRM returned irrelevant or inconsistent results
  • Consultants wasted time re-sourcing candidates already in the system

 

  1. Pipeline blind spots

 

  • Leaders couldn’t answer simple questions:
    • How many candidates do we have for X skillset in Y region?
    • What’s our real job coverage?
    • Where are we losing candidates in the funnel?

 

  1. Automation failing due to inconsistent data

 

Automations were built on assumptions that didn’t match real data usage:

  • Missing fields → broken workflows
  • Inconsistent statuses → no engagement
  • Poor segmentation → irrelevant messaging

 

  1. Untapped revenue locked in legacy data

 

Great candidates existed in the system — but couldn’t be found, reactivated or matched effectively.

The company didn’t have a technology director internally, and regional operations lacked alignment.

OUR APPROACH

  1. Data Audit & Classification

 

We began with a full ecosystem analysis, identifying:

  • The true state of data hygiene
  • Broken or redundant fields
  • Duplicate density
  • High-value segments buried in the system
  • Gaps preventing accurate search and automation

 

We produced a red/amber/green assessment covering all core objects.

 

  1. Data Cleaning & Normalisation

 

We rebuilt the data structure to make Bullhorn a usable search engine again:

  • Merged duplicates at scale
  • Normalised core fields (skills, locations, statuses, categories)
  • Archived non-usable legacy records
  • Rebuilt skills taxonomy
  • Introduced validation rules
  • Applied naming and formatting consistency

 

This created a clean, trustworthy foundation for sourcing, automation and AI.

 

  1. Re-Engineering Sourcing Workflows

 

We redesigned workflow stages so consultants could:

  • Run consistent searches
  • Find candidates based on skills, tags and experience
  • Auto-create shortlists from segmented data
  • Build talent pools by vertical and market

 

Recruiters immediately felt the difference.

 

  1. Activating the Data with Automation

 

Once the foundation was fixed, we deployed a full automation framework:

  • Candidate reactivation workflows
  • Skill-based nurture sequences
  • Market mapping sequences
  • Assignment follow-up
  • Compliance and screening flows
  • Job coverage prompts
  • Post-placement success workflows

 

These automations were personalised, brand-consistent, and written to reduce emotional variance across messaging.

 

  1. Converting Data Into Deal Flow

 

We built a repeatable cycle for turning dormant data into live revenue:

  • Reactivate existing candidates
  • Match to open roles
  • Auto-notify recruiters when a great candidate becomes available
  • Push real-time talent pools into BD conversations
  • Enable managers to forecast supply/demand by market

 

This created a direct link between data quality → sourcing efficiency → job delivery → revenue.

 

  1. AI-Readiness & Knowledge Layer

 

With a clean dataset, we prepared the CRM for AI-driven tools:

  • Structured, machine-readable fields
  • Standardised objects for embedding and retrieval
  • Organised notes and documents for summarisation
  • Removed clutter to reduce model confusion

 

The company is now in a position to adopt:

  • Recruiter copilots
  • Skill-based candidate matching
  • Automated shortlists
  • AI-powered BD insights
  • Semantic search and deep retrieval

THE IMPACT

  1. Recruiters Returned to the CRM

 

Bullhorn became the primary sourcing tool again.
Search accuracy increased dramatically.
Consultants stopped relying exclusively on LinkedIn.

 

  1. Reactivation Became a Revenue Stream

 

Automation resurfaced thousands of previously dormant candidates.
Several placements were generated within weeks from reactivation workflows alone.

 

  1. Fill Rates Increased

 

Better shortlisting + better job coverage = faster delivery.
Teams hit SLA milestones consistently for the first time.

 

  1. Talent Pools Became BD Fuel

 

Market maps generated directly from Bullhorn enabled:

  • More targeted BD conversations
  • Better client intelligence
  • Faster response to new vacancies

 

  1. Leadership Gained Real Visibility

 

Clean data unlocked:

  • Funnel health
  • Market heat mapping
  • Forecast accuracy
  • Quality-of-pipeline reporting

 

Decision-making became data-led rather than intuition-led.

 

  1. AI Adoption Roadmap Activated

 

With structured, reliable data, the agency is now rolling out:

  • AI sourcing assistants
  • Candidate summarisation
  • Semantic search
  • Automated shortlist generation
  • Client intelligence tools

 

All powered by the clean, unified Bullhorn ecosystem we created.

 

What This Enabled

 

The transformation didn’t just improve data quality — it changed the trajectory of the business:

  • Faster delivery
  • Higher consultant productivity
  • Consistent international processes
  • Less admin
  • More placements from existing data
  • A scalable, AI-ready CRM environment

 

The company now treats Bullhorn as a strategic asset, not an administrative burden.

CLIENT QUOTE

“We always assumed our data was ‘too messy’ to fix. Minima Maxima proved otherwise — within months we were generating real revenue from segments of the database we didn’t even know we had.”

Scaling a 40-Person Recruitment Agency with a Unified  Ecosystem

OVERVIEW

A fast-growing, sector-focused recruitment agency with 40 staff was struggling to scale due to fragmented systems, inconsistent workflows and poor data foundations.
Despite strong market demand, the business found it increasingly difficult to maintain delivery quality, forecast accurately, and onboard new consultants effectively.

Minima Maxima partnered with the agency to modernise its entire  ecosystem — aligning people, process, data and technology into one unified, scalable operating model.

THE CHALLENGE

The agency faced several common but serious obstacles to growth:

 

  1. Disconnected Tools & Manual Work

 

  •  CRM used differently across teams
  • Ad-hoc spreadsheets driving core processes
  • Manual compliance steps and duplicated admin
  • Automations not aligned to real recruiter workflow

 

  1. Poor Data Quality

 

  • Duplicate candidates and contacts
  • Inconsistent job and pipeline usage
  • Missing critical data breaking reporting

 

  1. Lack of Insight

 

  • Leaders unable to see reliable revenue/GP forecasting
  • No visibility of job coverage, pipeline health or leakage
  • Reporting inconsistent across divisions

 

  1. Scaling Bottlenecks

 

  • New hires struggled with onboarding
  • Processes varied widely between consultants
  • Tech stack not supporting the pace of growth

 

The business knew there CRM could and should do more — but lacked the internal bandwidth and expertise to fix, redesign, and embed it properly.

OUR APPROACH

  1. Discovery & Mapping

 

We conducted a structured assessment of:

  • Current  usage
  • Recruiter workflow (real behaviour vs intended process)
  • Data health
  • Automations
  • Reporting needs
  • Team adoption and friction points

 

This gave a 360° view of how the agency actually operated across front, middle and back office.

 

  1. Redesigning the Ecosystem

 

We rebuilt the CRM to match the agency’s real-world operations:

  • Cleaned and standardised all pipelines
  • Introduced meaningful job stages and required fields
  • Built structured candidate workflows
  • Re-aligned recruiter journeys with automation logic
  • Standardised data taxonomy across teams
  • Created a single, documented way of working

 

This gave consultants clarity, consistency and a more intuitive system.

 

  1. Data Hygiene & Governance

 

We created a “data-ready-for-AI” environment by:

  • Removing duplicates
  • Fixing broken fields
  • Establishing validation rules
  • Introducing data governance rhythms
  • Giving managers visibility of data quality KPIs

 

This tackled the root causes of poor reporting and workflow friction.

 

  1. Automation Framework

 

We designed and implemented a full-service automation layer, including:

  • Job coverage and follow-ups
  • Candidate screening and availability checks
  • Compliance reminders
  • Nurture sequences
  • Post-placement care
  • Internal ops workflows

 

Automations were written to reinforce brand tone and reduce emotional variance across communications.

 

  1. Analytics & Insights Layer

 

We built out there Analytics capability that gave leaders real visibility:

  • Job coverage
  • Submissions metrics
  • Funnel leakage
  • Contractor book performance
  • Desk productivity
  • Forecasting dashboards

 

This enabled data-led decision-making for the first time.

 

  1. Training & Embedding

 

We delivered:

  • Team training sessions
  • Manager toolkits
  • Adoption playbooks
  • Video walkthroughs
  • Onboarding packs for new hires

 

The result: the system stuck.

THE IMPACT

  1. Job Coverage Increased

 

Recruiters had clearer workflows and automations prompting key actions.
This led to:

  • More jobs being worked
  • Faster submission times
  • Better candidate engagement

 

  1. Reporting Became Trustworthy

 

Leaders finally had:

  • Accurate revenue forecasts
  • Real-time desk performance
  • Visibility across all divisions

 

This improved planning, hiring and resource allocation.

 

  1. Recruiter Output Increased

 

By removing manual admin and smoothing the workflow, consultants gained approx:
+5–7 hours per week of productive time
which they redirected into sourcing delivery and further training.

 

  1. Reduced Operational Risk

Compliance steps were automated, documentation was standardised, and data governance improved — significantly reducing risk.

 

  1. System Adoption Unified

Every consultant now used the CRM the same way, enabling the business to scale consistently across teams and new hires.

 

  1. Higher Margin & Better Utilisation

With clearer visibility and fewer bottlenecks, the agency saw:

  • Improved fill rates
  • More consistent pipelines
  • Stronger utilisation across teams

 

What This Enabled

 

The agency emerged with a scalable, modern CRM ecosystem — not just a tidied-up system, but a platform capable of supporting their next stage of growth:

  • Faster onboarding
  • Predictable revenue
  • Stronger delivery
  • AI-ready data foundations
  • Streamlined operations
  • A more consistent client and candidate experience

CLIENT QUOTE

“Minima Maxima helped us transform  from a system we used into a system that drives the business.
Our workflows are cleaner, reporting is accurate, and our consultants are noticeably more productive.”

Complete the form below to see and download the AI-Ready Data Checklist.