Machine Learning Engineer

Sayed Raheel Hussain

I design and ship production-grade AI systems under real-world constraints. My work focuses on voice-first LLM pipelines, multi-agent systems, and applied computer vision for healthcare and logistics — with full ownership from data to deployment.

LinkedIn GitHub Explore Work

3+ yrs ML Systems

$6M in 8 months ARR Impact

50+ Ops Scale

Voice AI • LLM Orchestration • Computer Vision

010101 110010 011001 101010 • SYSTEMS THAT SCALE IN PRODUCTION

Achievements & Recognition

Business Insider

Recognized for applied innovation in AI-driven logistics at Kayaan AI.

01000010 01001001

NVIDIA Inception

Selected for NVIDIA Inception program for AI product scale.

01001110 01010110

Google for Startups

$100K awarded to accelerate applied AI systems.

01000111 01001111 01001111 01000111 01001100 01000101

Microsoft for Startups

$150K awarded for ML deployment and growth.

01001101 01001001 01000011 01010010 01001111

$1M ARR

0→1 AI systems reached $1M ARR in under 8 months.

01000001 01010010 01010010

Signature Projects

Deep learning systems that are built to be shipped, monitored, and iterated.

Cow Health Analyzer

End-to-end mastitis detection platform for dairy farms.

Vision • Healthcare

PathoAgent VQA

Multimodal pathology assistant built on ViLT.

Multimodal • Education

MedicoBot

RAG-powered medical Q&A bot using LLAMA 2, LangChain, FAISS.

GenAI • RAG

Lung Segmentation

SegNet + ResNet-101 with 81.5 IOU and 93 Dice score.

Medical • Segmentation

Experience

Kayaan AI — Machine Learning Engineer

Jan 2025 – Present | New York, NY

Built a voice-driven logistics system (ASR → LLM → TTS) with evaluation pipelines and safety guardrails. Automated 30–40% workflows and scaled ops from ~5 to ~50 concurrent workflows.

Sync — Applied ML Engineer

Sep 2024 – Jan 2025

Built and deployed multi-agent AI assistant for clinical information retrieval.

Palazzo.ai — Machine Learning Engineer

Jan 2024 – Jul 2024 | Philadelphia, PA

Fine-tuned ControlNet diffusion models and productionized segmentation + generation pipelines, improving throughput by 30%.

Yeshiva University — Research Assistant

Jun 2023 – May 2024

Built VetMedGPT with 500GB+ corpus and 1B-parameter transformer on 6× A100 GPUs. Fine-tuned YOLOv8 segmentation to 54%+ AP.

Research & Open Source

VetMedGPT (2024)

Domain-specific transformer for veterinary healthcare and clinical decision support.

BovineTeatNet (2024)

Few-shot health assessment with YOLOv8 segmentation on limited labeled data.

LlamaSearch

Multi-agent system that routes search across Google, ArXiv, and Scholar.

Let’s Build Something Bold

Open to ML engineering roles, research collaborations, and applied AI startups.

Connect on LinkedIn See GitHub