Real-world automation solutions delivering measurable business impact across diverse industries
An investment firm needed real-time access to comprehensive UK company data from Companies House for due diligence, competitive analysis, and investment opportunity identification. Manual data retrieval from the Companies House website is inefficient, error-prone, and can't scale to analyze thousands of companies.
Engineered a web scraping and data aggregation tool that continuously monitors Companies House records. The system extracts company financials, director information, filing history, shareholder structures, and legal proceedings. Data is normalized, enriched with additional public sources, and delivered through a searchable dashboard with real-time alerts for significant changes (directorship changes, financial filings, insolvency notices).
Automated monitoring of 50,000+ UK companies
User-friendly, well presented data dashboard for the investors
99.7% data accuracy rate
Early identification of 4 high-value investment opportunities
Building a professional portfolio website from scratch requires 3-6 hours of manual data entry, design decisions, and content writing — time and skills that most job seekers, students, and freelancers simply do not have.
Built Foliage, an AI-powered builder that uses GPT-4o to parse an uploaded resume and instantly generate a fully structured portfolio. Users pick from 10 professionally designed templates in a live gallery, then customize every section in a three-panel editor with real-time preview. The finished site downloads as a fully self-contained HTML file or saves to cloud storage via Supabase. No coding, no design skills, no manual data entry.
Resume to finished portfolio in under 20 minutes
90-95% reduction in time vs. manual portfolio building
10 professionally designed templates with live preview
Fully self-contained HTML export — no hosting required
Multi-resume support with instant AI re-generation
The tech transfer office struggled to effectively market diverse university innovations to appropriate industry partners, investors, and potential licensees. Each technology required customized outreach campaigns, but manual segmentation and personalized messaging was overwhelming the small marketing team. Response rates were low and many promising technologies received insufficient market exposure.
Built an AI-powered marketing automation platform that analyzes each new technology disclosure and automatically generates multi-channel marketing campaigns. The system uses natural language processing to understand the innovation's technical details and market potential, then creates targeted campaigns including email sequences, social media content, industry-specific landing pages, and personalized outreach to relevant contacts in the CRM. Machine learning algorithms continuously optimize messaging and targeting based on engagement data.
over 300% increase in response inquiries
30 minutes saved per marketing campaign
Automated outreach to 1000+ industry partners
A spring-clip mechanism for an EUV reticle needed to survive real operational loads while staying precise to within microns. Manual multi-physics analysis across materials and load cases would have taken days.
Ran AI-driven structural, strain, and compliance analysis across PETG, PEEK, and Al 6061-T6. Compressed days of analysis into hours, identified the dominant failure point, and derived a hard 0.83mm pad thickness limit for sub-5μm displacement precision.
Dominant failure point identified in hours vs. days
Resolved stiffness vs. actuation stroke design conflict
Derived hard 0.83mm pad thickness limit for sub-5μm precision
Revealed pad shear deformation and frictional slip as independent mechanisms
Validated t³ stiffness scaling law across all geometries
The University of Delaware Tech Transfer Office needed to increase visibility of emerging technologies and innovations available for licensing. Manual post generation to LinkedIn was time-consuming, inconsistent, and failed to reach optimal posting times for maximum engagement.
Developed an intelligent LinkedIn automation system that automatically monitors new technology disclosures from the university's innovation pipeline. The system generates compelling, professional posts with relevant hashtags, images, and descriptions tailored to different audience segments (investors, industry partners, researchers). Posts are scheduled at optimal times based on engagement analytics.
2 hours saved weekly on social media management