We build AI systems that work in the real world.
From supervised and unsupervised machine learning to enterprise agentic workflows, we design, deploy, and scale intelligent solutions for modern businesses.
Want to talk through your product idea, roadmap, or build plan? Book time directly on Franklin's Calendar and pick a slot that works for you.
Built a lightweight CNN to detect malaria from microscopic blood smear images. After testing baseline CNNs, transfer learning, HSV exploration, and augmented models, the final domain-specific model reached approximately 98%+ test accuracy.
Built a real-world photo verification platform for authentic image capture. Verisnap combines trusted camera capture, device-signal analysis, TruthScore-style verification, enterprise workflows, and structured reporting for teams that need reliable visual evidence.
Built an AI-powered pipeline that ingests scans, images, HTML, and PDFs, runs OCR and NLP extraction, maps clinical entities to ICD, CPT, and RxNorm codes, and creates a searchable index for downstream medical record review.
Built an end-to-end AI workflow that converts PDFs into domain-specific Q&A datasets, fine-tunes an open-source LLM with LoRA, merges the model, and exports it to GGUF for lightweight local inference.
Half Heart Intelligence is led by Franklin Perez, a veteran software engineer, founder, and technology leader with roughly 35 years of software development experience across startups, backend systems, high-performance platforms, and product delivery.
We donate 3% of our proceeds to Hope for Heart Kids.