Systems Portfolio

Intelligent Systems

A curated collection of AI tools, automation frameworks, and product systems. Each represents a real problem solved with architecture and intent.

NovaBridge Capital: SME Credit Transformation

Live

Problem

Manual credit assessment processes caused 3-week approval cycles, high analyst workload, and inconsistent risk decisions across markets.

Outcome

Reduced loan approval cycle from 21 days to 4 days. Standardised risk decisioning across 6 markets. Delivered in 9 months.

21 days to 4 days6 markets standardised9-month delivery
Power AppsAzureML ModelsSharePointPower BI
Read more →

Zenbridge Bank: KYC Operations Automation

Live

Problem

KYC operations were entirely manual, causing backlogs, compliance gaps, and high operational overhead ahead of a CBN audit.

Outcome

80% reduction in manual KYC processing time. Full CBN compliance achieved. Operational team redeployed to higher-value tasks.

80% processing time reductionCBN compliant7-month delivery
Power AutomateSharePointMicrosoft 365Power BI
Read more →

Stanbic IBTC: EasyBuy Consumer Finance

Live

Problem

Customers lacked accessible financing options at point of purchase, limiting retail partner conversion and bank revenue.

Outcome

Launched EasyBuy across multiple retail verticals with streamlined onboarding and instant POS credit decisioning.

Multi-vertical launchInstant POS creditRetail partner network
Product StrategyAgile DeliveryAPI IntegrationCRM
Read more →

Learning & Development Programme Design

Live

Problem

Teams lacked the skills to operate newly deployed systems, risking adoption failure and productivity loss post-implementation.

Outcome

High adoption rates achieved across trained cohorts. Measurable improvement in process compliance and system utilisation within 60 days.

90% error reductionApproval cycle < 24 hrs95% adoption in 3 months
Power AppsPower AutomateSharePointPower BI
Read more →

AFI: Online Food Ordering & Reservation System

Live

Problem

Manual and fragmented food ordering and reservation processes led to customer dissatisfaction, operational bottlenecks, and zero real-time visibility for managers.

Outcome

Delivered high-fidelity Figma prototypes across four user roles with validated journey maps, empathy mapping, and impact-complexity prioritisation.

4 user role prototypesFull journey mappingWCAG compliant design
FigmaReact.jsNode.jsMongoDBAWSRESTful APIs
Read more →

AI Resume Analyzer

Live

Problem

Job seekers struggle to understand why their resumes are rejected by ATS systems.

Outcome

3x improvement in interview callback rate for beta users.

94% ATS accuracy3x callback rate500+ analyses run
Next.jsClaude APITypeScriptSupabase

AI Research Agent

Live

Problem

Research tasks consume 60% of analyst bandwidth with diminishing accuracy at scale.

Outcome

Delivers structured research reports in under 3 minutes.

< 3 min reports85% analyst time saved12 domains covered
LangChainClaude APINext.jsTavilyTypeScript

DigiAssist: Stanbic IBTC Mobile Onboarding

Live

Problem

Branch-dependent account opening created long queues and a 24-hour turnaround time, limiting reach and frustrating both customers and staff.

Outcome

Reduced account opening TAT from 24 hours to 5 minutes. Enabled fully mobile, field-based account creation for Tier 1 and Tier 3 accounts. Delivered within 3 months.

24 hrs to 5 mins TATTier 1 & 3 accounts3-month delivery
Mobile AppProduct ManagementAgile DeliveryAPI IntegrationUX Design

Multi-Agent Market Research System

Live

Problem

Smart Money Nigeria's manual research and reporting process for its professional investor community limited the speed, consistency, and scalability of market intelligence delivery.

Outcome

Automated the full research pipeline from data gathering to structured report generation, enabling faster, data-driven market insights at scale for the investment community.

Automated end-to-end researchStructured report outputScalable insight delivery
Multi-Agent AIClaude APILangChainWeb Scrapingn8nTypeScript
Read more →

Automation Workflow Engine

Live

Problem

Repetitive manual workflows across tools drain team productivity.

Outcome

Automated 15 workflows saving 20 hours/week per client.

20 hrs/week saved15 workflows automated4 client deployments
n8nNode.jsPostgreSQLREST APIsDocker