Skip to main content
Download free report
Softblues
Softblues
Our Work

AI Case Studies

46 projects, from AI agents and voice assistants to document automation and internal copilots. Each one lists the problem, what we built and the numbers that came out of it.

Flagship AI projects

The 8 projects we point to most often.

AI-Powered Candidate Assessment Platform
HR & RecruitmentAI Business Automation

SofiaHR

AI-Powered Candidate Assessment Platform

How Softblues automated SofiaHR's candidate-assessment process: a 2 to 3 hour expert linguistic analysis, scored across 10 psychological dimensions, now runs in minutes at 90%+ of human-analyst accuracy.

2-3 hrs → mins
Per-Candidate Analysis
90%+
Accuracy vs Expert Analysts
Fine-tuned Gemini Flash (10 models)Claude SonnetVertex AIOpenAI Realtime APIElevenLabs (STT, diarization, TTS)+10 more
View Case Study
AI-Powered Pharmacy Operations Assistant
HealthcareAI Business Automation

ailiRX

AI-Powered Pharmacy Operations Assistant

How Softblues designed and validated an eight-agent automated pipeline for pharmacy operations: the 15 to 20 minute, six-step prescription workflow behind every script, compressed toward seconds without dropping a compliance check.

15–20min → secs
Per-Prescription Workflow
8
Specialised AI Agents
Gemini 2.5 ProMulti-Agent OrchestrationPythonFastAPIGoogle Cloud Platform+11 more
View Case Study
AI-Powered Clinical Research Platform
HealthcareAI Business Automation

Lumono.ai

AI-Powered Clinical Research Platform

How Softblues automated a 12 to 18 month manual research process into a nine-agent AI pipeline that turns a plain-English question into a publication-ready analysis in weeks.

Months→Weeks
Research Cycle
9
AI Agents Orchestrated
Claude SonnetLangChainMulti-Agent OrchestrationPythonFastAPI+12 more
View Case Study
The Softblues Claude Operating System
SoftbluesAI Business Automation

Running our own company on a network of AI agents

The Softblues Claude Operating System

Softblues runs its own company on Claude: six connected agents, one per function, each with its own operating manual, tools and routines, joined through a shared hub. We use it before we sell it, so we can show it live rather than in a slide deck.

6
Connected Spaces
3+
Functions Run by Agents
Claude (Anthropic)CoworkSkillsScheduled TasksStatus Briefs+9 more
View Case Study
From Missed Calls to 24/7 Support
Customer SupportAI Business Automation

Roof Maker

From Missed Calls to 24/7 Support

Roof Maker's support line went quiet after hours, and routine questions ate up specialist time during the day. Softblues automated the front line of their customer support: an AI voice agent that answers every call, troubleshoots, logs the ticket in HubSpot, and escalates anything urgent to a person. 24/7, in production.

24/7
Support Coverage
8
AI Agents
PythonLiveKitDeepgram Nova 3OpenAI GPT-4.1-miniElevenLabs Flash v2+7 more
View Case Study
Order-to-Schedule Automation for a Secure Logistics Operator
Secure LogisticsAI Business Automation

Cash-in-transit operator (client anonymised)

Order-to-Schedule Automation for a Secure Logistics Operator

A cash-in-transit operator ran its scheduling on email and spreadsheets: hundreds of order emails a day, hand-built run sheets, and one person's knowledge holding it together. We ran a four-week discovery, mapped the whole process and its 70+ rules, and designed an order-to-schedule automation that runs inside the client's own Microsoft tenant.

15–30 hrs/wk
Scheduling Time to Reclaim
70+
Business Rules Mapped
Claude SonnetClaude OpusClaude Vision (OCR)PythonLangGraph+9 more
View Case Study
Automating the Monthly Compliance File Review
Financial ServicesAI Business Automation

Regulated financial-advice firm (client anonymised)

Automating the Monthly Compliance File Review

A regulated financial-advice firm reviews every client advice file by hand each month: dozens of mechanical checks per file plus a judgement call on suitability, all under a regulator's eye. We proposed a four-stage multi-agent pipeline that does the mechanical work and drafts the review, with a compliance reviewer keeping the final sign-off. It is designed to run entirely inside the firm's own Microsoft tenant, in the EU.

~40
Checks per File
4
Pipeline Stages
Claude SonnetClaude OpusGPT-5.5Microsoft Foundry Agent ServiceMicrosoft Agent Framework (Python)+6 more
View Case Study
Auditing an AI-Built App and Making It Safe to Scale
Food ProductionClaude Code Setup

Claude Code audit & setup, food producer (client anonymised)

Auditing an AI-Built App and Making It Safe to Scale

A food producer's own team built a real operational app with AI: procurement, stock and traceability, and it works. Underneath, it was a 58,000-line monolith with critical security holes and no tests, weeks from go-live. We ran a Claude Code audit and discovery, closed the security exposure, and set out a secure, modular rebuild plus an AI development setup the team keeps building on.

58K
Lines Audited
4
Priority Findings
Claude CodeClaudeFirebaseReactTypeScript+7 more
View Case Study

The full back catalogue

38 further projects, grouped by industry.

Want the same for your process?

Bring us one workflow that costs you time every week. We will tell you whether AI is the right answer and what it would take.