Available for Senior AI/ML Roles — Open to Relocation

Currently Building: Bio-Oracle v2.0

Building AI That
Ships to Production

From neuro-symbolic agents and enterprise RAG pipelines to edge-deployed computer vision — I engineer AI systems that work beyond the notebook.

Currently building:

View Projects Resume harshrofff@gmail.com

Efficiency Gain
VITG Production RAG

CV Accuracy
Edge Deployment

Docs Automated
Regulatory PDFs

Publications
IEEE + Elsevier

// about

Who I Am

I'm an AI/ML Engineer who specializes in designing and deploying production-grade AI systems for regulated industries. My work spans architecting multi-agent LLM pipelines on AWS Bedrock, building neuro-symbolic reasoning systems, and shipping computer vision to edge hardware.

MS in Data Science from UMBC (3.8 GPA), with active research under US Army Research Lab funding in distributed sensing and autonomous systems. I've co-authored two peer-reviewed papers (IEEE + Elsevier), hold an AWS Machine Learning Associate certification, and care deeply about AI safety, auditability, and production observability.

I believe the gap between "demo" and "deployed" is where real engineering happens. That's where I work.

MS Data Science

UMBC — GPA 3.8/4.0

AWS ML Certified

Active through Apr 2028

2 Peer-Reviewed Papers

IEEE ITU + Elsevier JAFR

Army Research Lab

Funded Research, UMBC CRDSA

// skills

Technical Stack

LLM Stack

LLM Agents RAG Systems Neuro-Symbolic AI PydanticAI LangChain Prompt Engineering RLHF Fine-tuning Multi-model Orchestration

Core AI / ML

PyTorch TensorFlow Computer Vision YOLOv8 OpenCV Transformers CNNs / RNNs Anomaly Detection Time Series

Cloud & MLOps

AWS Bedrock SageMaker Lambda Textract MLflow W&B Docker CI/CD Model Governance A/B Testing

Data Engineering

Spark Hadoop PostgreSQL MongoDB Redis Pinecone pgvector ETL Pipelines Feature Engineering

Software & Deploy

Python FastAPI REST APIs Streamlit Gradio CUDA NVIDIA Jetson Edge Computing Raspberry Pi

// featured

Production AI Systems

End-to-end engineering — not tutorials.

NEURO-SYMBOLIC AI

Bio-Oracle

A neuro-symbolic AI agent for high-throughput drug discovery screening. Combines deep learning segmentation (Cellpose) with LLM-based structured reasoning (PydanticAI + Gemini) to automate phenotypic analysis on the BBBC021 dataset — where pure deep learning fails interpretability requirements.

Problem Drug discovery screening requires CV precision + auditable reasoning — pure DL fails interpretability.

Approach Neuro-symbolic agent: Cellpose segmentation → 30+ morphological features → PydanticAI + Gemini reasoning.

Outcome Verifiable reasoning chains on BBBC021 dataset; optimized for Apple Silicon MPS local inference.

Cellpose Segmentation Feature Factory (30+ features) Robust Z-score Norm PydanticAI Agent Gemini Reasoning

PydanticAI Gemini Cellpose Docker Apple MPS

View Code

CONVERSATIONAL AI

PelloHorter

AI-powered phone agent conducting autonomous outbound calls with human-like conversational flow.

Problem Manual candidate screening calls don't scale — repetitive, time-consuming, inconsistent.

Approach Real-time voice synthesis + LLM dialogue management + dynamic response generation pipeline.

Outcome Fully autonomous screening and qualification workflows operating at scale.

Voice AI LLM Speech Processing Real-time Inference

Private Repo

RAG SYSTEM · PRODUCTION

Enterprise Document Intelligence

Production RAG system automating 2,500+ regulatory PDFs with SOC2-compliant audit trails.

Problem Manual compliance review of 2,500+ unstructured regulatory PDFs was a critical bottleneck.

Approach AWS Textract OCR → pgvector embeddings → Claude 3.5 on Bedrock → Streamlit audit dashboard.

Outcome 60% cut in screening time, 40% efficiency gain, 100% auditability, SOC2-compliant.

AWS Bedrock SageMaker Textract pgvector Streamlit

Proprietary — VITG Corp

CV · PUBLISHED

FilletCam AI

Problem Manual fish fillet grading is slow and inconsistent on production lines.

Approach YOLOv8 + OpenCV deployed on Raspberry Pi for real-time edge inference.

Outcome 92% accuracy. Published in Elsevier JAFR 2024.

YOLOv8 OpenCV Raspberry Pi

Elsevier JAFR

EDGE AI · PUBLISHED

Mosquito Wing Beat Classification

Problem Species-level mosquito ID needed in resource-constrained field environments.

Approach TinyML audio classifier on ARM Cortex-M — no network, no GPU.

Outcome <100ms inference. Published IEEE ITU Kaleidoscope 2021.

ARM Cortex-M TinyML Python

IEEE ITU

AI PLATFORM · LIVE DEMO

Silicon Oracle

Full-stack AI stock analysis & paper trading platform. BYOK architecture — each user supplies their own API keys, so rate limits never hit everyone at once. Real-time market data, Oracle Score™ (15-factor analysis), Gemini-powered email intelligence, and Alpaca paper trading.

Problem Retail investors lack professional-grade AI analysis without expensive subscriptions or shared rate-limit bottlenecks.

Approach BYOK Flask platform: Finnhub market data → Oracle Score™ (15-factor system) → Gemini 2.0 Flash AI analysis → Alpaca paper trading → automated hourly email intelligence alerts.

Outcome Live multi-user platform with real-time analysis, hourly AI market intelligence, paper trading, and portfolio tracking — deployed free on Render.

Flask Gemini 2.0 Flash PostgreSQL Alpaca API Supabase Tailwind CSS APScheduler

Live Demo Private Repo

// projects

More Work

AI OpenAI

Multi-Modal Recommendation Engine

Hybrid recommendation system combining structured data, computer vision, and LLM personalization. A/B tested for continuous improvement.

−20% bounce • +40% engagement

Python · OpenAI API · PostgreSQL · A/B Testing

Data GeoSpatial

Infrastructure Risk Assessment

Real-time geospatial risk analytics for transportation infrastructure. Multi-source data fusion, interactive Leaflet maps, PostgreSQL/PostGIS backend, color-coded risk overlays for decision support.

Flask · PostgreSQL · PostGIS · Leaflet

Data ML

Real-Time Analytics Dashboard

Distributed ML system for multi-modal sensor data analytics. Real-time inference, interactive Plotly visualizations, 20% performance improvement on optimization tasks (Army Research Lab).

+20% performance

Python · Dash · Plotly · PyTorch

CV Edge

Gesture-Based Control System

Real-time gesture recognition for robotic system control. Deployed on NVIDIA Jetson with CUDA-optimized inference for low-latency embedded operation in field environments.

Python · OpenCV · NVIDIA Jetson · CUDA

// experience

Where I've Worked

Mar 2023 — Present Current

AI/ML Researcher

UMBC Center for Real-time Distributed Sensing and Autonomy

Army Research Lab funding — built distributed ML systems for multi-modal sensor fusion; 20% performance improvement in real-time optimization
Designed and deployed CHARLIE, a voice-enabled AI agent (speech recognition + LLM inference + real-time dialogue) for autonomous system control — Demo ↗
Engineered production CV systems on NVIDIA Jetson edge devices for real-time perception pipelines

PyTorch TensorFlow NVIDIA Jetson Voice AI LLM

Jun 2024 — May 2025

AI/ML Engineer — Production AI Systems

VITG Corp., Halethorpe MD

Architected production RAG system on AWS Bedrock (Claude 3.5) — processed 2,500+ regulatory PDFs, cut screening time by 60%
Built semantic search with vector embeddings + AWS Textract — 100% auditability, SOC2-compliant logging
Established ML governance frameworks: monitoring pipelines, hallucination evaluation metrics, LLM observability dashboards
Built scalable ETL pipelines + FastAPI services for multi-model LLM orchestration in production

AWS Bedrock SageMaker FastAPI Docker PostgreSQL

Aug 2023 — Dec 2023

Data Scientist Intern — Computer Vision

The Conservation Fund, Shepherdstown WV

Built YOLOv8 + OpenCV pipeline achieving 92% accuracy — deployed on Raspberry Pi for real-time edge quality assessment
Co-authored peer-reviewed research; published in Elsevier Journal of Agriculture and Food Research (2024)

YOLOv8 OpenCV Raspberry Pi Edge ML

May 2021 — Jun 2022

ML Engineer — Applied AI & Audio Analytics

WeHear Innovations Pvt. Ltd., Ahmedabad India

Developed Personal Hearing Intelligence (PHI) ML models — longitudinal audio analysis for personalized hearing-risk metrics on bone-conduction hardware
Built time-series feature engineering pipelines for high-frequency sensor data; early-risk detection through statistical monitoring
Integrated model outputs into production mobile apps via firmware team collaboration, translating raw audio predictions into clinical insights

Python Time Series Audio Processing Mobile Integration

Master of Science, Data Science

University of Maryland Baltimore County (UMBC)

Aug 2022 – May 2024 • GPA: 3.8 / 4.0

ML · Deep Learning · NLP · Computer Vision · Big Data Systems · Statistical Analysis

// research

Publications & Credentials

Peer-Reviewed Publications

Elsevier · Journal of Agriculture and Food Research · 2024

FilletCam AI: Precision color profiling of fish fillets using deep learning

Ranjan, R., Shroff, H., et al.

YOLOv8 + OpenCV pipeline for automated fish fillet quality grading deployed on edge devices, achieving 92% accuracy. Enabled commercial AI quality control systems.

DOI: 10.1016/j.jafr.2024.101461

IEEE · ITU Kaleidoscope Conference · 2021

Mosquito identification using machine learning on embedded systems

Trivedi, K., Shroff, H.

TinyML pipeline for real-time mosquito wing beat classification on ARM Cortex-M microcontrollers with sub-100ms inference latency for field deployment.

DOI: 10.23919/ITUK53220.2021.9662116

Certifications

Machine Learning Associate

Amazon Web Services · Active through Apr 2028

Machine Learning Foundations

Amazon Web Services

Transformer-Based NLP Applications

NVIDIA Deep Learning Institute

GPU-Accelerated Computing