AI Research Engineer · IEEE Published · MS Computer Science

NikhileshVerma.

LLM Systems · Edge AI Security · Spatiotemporal ML · Autonomous Intelligence

Graduate Research Assistant — Texas A&M University–Corpus Christi
Software & System Security Lab · CAHSI–Google Scholar

IEEE-published researcher building production-grade AI at the intersection of large language models, retrieval-augmented generation, and on-device AI security. Sub-100ms LLM inference, multi-enclave TEE on RISC-V, and terabyte-scale distributed ML pipelines. 8+ open-source repositories.

0
IEEE Papers
0
GPA / 10
0
OSS Repos
0
Hallucination ↓
Nikhilesh Verma — AI Research Engineer
◈ IEEE UPCON 2025
⚡ SafeNav-RAG
🔒 RISC-V TEE
Scroll to explore
GPT-4 · Claude · Gemini · LLaMA · Phi-4 LangChain · LlamaIndex · LangGraph FAISS · BM25 · Hybrid RAG · pgvector PyTorch · TensorFlow · Hugging Face FastAPI · Docker · Kubernetes · HPC/GCP RISC-V · Keystone TEE · Edge AI · SiFive CNN-LSTM · Transformers · Autoencoders · SHAP Multi-Agent Systems · Agentic Workflows GPT-4 · Claude · Gemini · LLaMA · Phi-4 LangChain · LlamaIndex · LangGraph FAISS · BM25 · Hybrid RAG · pgvector PyTorch · TensorFlow · Hugging Face FastAPI · Docker · Kubernetes · HPC/GCP RISC-V · Keystone TEE · Edge AI · SiFive CNN-LSTM · Transformers · Autoencoders · SHAP Multi-Agent Systems · Agentic Workflows
Research Impact
0
IEEE Publications
UPCON 2025 (Published)
IEEE Big Data (Under Review)
0
Open-Source Repos
LLM · RAG · Edge AI · Multimodal
0
AI Research Domains
Edge Security · RAG · Spatiotemporal ML
0
Hallucination Reduction
SafeNav-RAG · Sub-100ms latency
Currently Open To
Target Roles
AI Research EngineerML Engineer Research ScientistGenAI Engineer PhD PositionsResearch Collaborations
Quick Facts
Corpus Christi, TX — Open to Relocation & Remote
IEEE-published in AV safety AI & crypto forecasting
CAHSI–Google sponsored research on RISC-V edge AI security
Production LLM systems with sub-100ms inference latency
8+ open-source AI repositories on GitHub
Research01/09

Research Work

🛡️
CAHSI–Google · Completed

On-Device AI Security — Multi-Enclave TEE on RISC-V

Designed and implemented multi-enclave Trusted Execution Environments on RISC-V (SiFive Unmatched, Keystone) to secure AI workloads on edge devices. Developed autoencoder-based anomaly detection to safeguard AI models against adversarial attacks, achieving a working prototype with multi-enclave isolation and ML-based anomaly detection on real hardware.

RISC-VKeystone TEESiFiveAutoencodersPyTorchC/C++Edge AI
🌩️
Active — CONUS Scale

CONUS Thunderstorm Prediction with Deep Learning

Expanding deep learning model from South Texas (400 km²) to full CONUS scale. Integrating NOAA HRRR, GOES-16 GLM, and NEXRAD radar datasets. Building terabyte-scale distributed ETL pipelines on HPC/GCP. Training CNN-LSTM architectures for spatiotemporal nowcasting across all U.S. climate zones with fault-tolerant data ingestion.

CNN-LSTMPyTorchNOAA HRRRGOES-16 GLMNEXRADHPC/GCPTensorFlow
🚗
Published · IEEE UPCON 2025

SafeNav-RAG: Latency-Aware RAG for Autonomous Vehicles

Production-grade low-latency RAG framework combining FAISS vector search and BM25 hybrid retrieval for real-time AV decision-making. Achieved ~55% reduction in LLM hallucination and sub-100ms inference latency. Microservice architecture deployed via FastAPI + Docker with CI/CD via GitHub Actions.

LangChainFAISSBM25FastAPIDockerEmbedding Models
📈
Under Review · IEEE Big Data

CryptoFusion: Temporal Graph-Aware Transformer

Multimodal forecasting architecture combining transformers, temporal graph neural networks (TGNs), and financial sentiment embeddings (FinBERT). Achieves ~8.3% RMSE reduction and +0.29 Sharpe ratio improvement over baseline financial models, demonstrating graph-aware temporal modeling in volatile crypto markets.

TransformersTGNFinBERTPyTorchSHAP
Architecture Diagrams02/09

System Designs

SafeNav-RAG Pipeline

AV Sensor Query Embedding Model FAISS BM25 Hybrid Reranker Top-K Context LLM Decision <100ms ~55% hallucination reduction · Sub-100ms · IEEE UPCON 2025

FAISS vector search + BM25 hybrid retrieval with reranking for safety-critical autonomous vehicle decision support.

RISC-V Multi-Enclave TEE

SiFive Unmatched — RISC-V Hardware Enclave 1 AI Model Weights Keystone TEE Autoencoder Monitor Enclave 2 Inference Engine PyTorch Runtime Anomaly Detection isolation Secure Monitor Attack Detection Adversarial Defense Secure Attestation CAHSI–Google · Multi-enclave isolation with ML-based anomaly detection

Multi-enclave TEE architecture on RISC-V with autoencoder anomaly detection for adversarially robust edge AI inference.

CONUS Thunderstorm Pipeline

NOAA HRRR GOES-16 GLM NEXRAD Radar Distributed ETL (HPC) Terabyte Scale CNN Spatial Features LSTM Temporal Seq Nowcast CONUS Scale Forecast South Texas 400km² → Full CONUS · Active Research · HPC/GCP

Terabyte-scale ETL pipelines from NOAA/NEXRAD/GOES-16 feeding CNN-LSTM spatial-temporal model for continental-scale nowcasting.

CryptoFusion Architecture

Price Data Graph Edges Sentiment Transformer TGN Encoder FinBERT Multimodal Fusion Cross-Attention Price Forecast 8.3% RMSE ↓ Portfolio Opt. +0.29 Sharpe SHAP Explain. Interpretability Transformer + TGN + FinBERT · IEEE Big Data · Under Review

Multimodal fusion of transformer price encoding, temporal graph networks, and financial sentiment for crypto forecasting.

Publications03/09

IEEE Publications

01✦ Published

SafeNav-RAG: Latency-Aware RAG for Autonomous Vehicle Decision-Making

Abhishek Joshi, Nikhilesh K. Verma, Alihan Hadimlioglu
Published at IEEE UPCON 2025 · IIT-BHU, Varanasi, India · 2025
~55%
Hallucination Reduction
<100ms
Inference Latency
Hybrid
FAISS + BM25
Real-Time
AV Decision Support

Production-grade low-latency Retrieval-Augmented Generation framework combining FAISS vector search, BM25 hybrid retrieval, and embedding pipelines for autonomous vehicle decision systems. Demonstrates ~55% reduction in LLM hallucination and sub-100ms inference latency, enabling real-time contextual decision support for safety-critical autonomous driving environments. Microservice-based inference deployed via FastAPI + Docker with CI/CD via GitHub Actions.

View on IEEE Xplore ↗
02◈ Under Review

CryptoFusion: Temporal Graph-Aware Transformer for Cryptocurrency Forecasting & Portfolio Optimization

Abhishek Joshi, Alihan Hadimlioglu, Nikhilesh K. Verma
Submitted to IEEE Big Data Conference · Under Review · 2025
~8.3%
RMSE Reduction
+0.29
Sharpe Ratio Gain
TGN
Temporal Graph Networks
Multi-modal
Sentiment + Price Fusion

Multimodal forecasting architecture combining transformers, temporal graph neural networks, and FinBERT sentiment embeddings for cryptocurrency price prediction and portfolio optimization. Achieves ~8.3% RMSE reduction and 0.29 Sharpe ratio improvement over baseline financial models, demonstrating the value of graph-aware temporal modeling in volatile, highly correlated crypto markets. SHAP analysis provides interpretability for trading decisions.

Achievements04/09

Awards & Achievements

📄
Research · Publication

IEEE UPCON 2025 — Paper Accepted & Published

SafeNav-RAG · IIT-BHU, Varanasi, India
2025
🏛️
Research Grant

CAHSI–Google Sponsored Research Project

Computing Alliance of Hispanic-Serving Institutions · Google Funded
2024–2025
🔒
Research

RISC-V Secure Edge AI — Working Prototype

Multi-enclave TEE + Autoencoder Anomaly Detection · SiFive Unmatched
2024–2025
🌩️
Research

CONUS Thunderstorm Prediction System

400 km² South TX → Full U.S. CONUS Scale · CNN-LSTM · Active
2025 — Ongoing
Academic Excellence

9.5 / 10 GPA — B.E. Computer Engineering

G.V. Acharya Institute · Mumbai University
2022
🎓
Graduate Research

Graduate Research Assistant — AI Security Lab

Texas A&M University – Corpus Christi · Software & System Security Lab
Jan 2025 – Present
💻
Open Source

8+ Open-Source AI Repositories

LLM · RAG · Edge AI · Multimodal · CLIP · FastAPI · RISC-V
github.com/nikhileshkverma
🤝
Leadership

VP — International Student Organization

TAMUCC · Supported 200+ international students through mentoring & advocacy
2025 – Present
🏆
Leadership

President — Computer Engineering Student Assoc.

CESA · Hackathons, workshops for 150+ students · 40% participation increase
2021–2022
Conferences & Talks05/09

Conference Presentations

2025

SafeNav-RAG: Latency-Aware RAG for Autonomous Vehicle Decision-Making

IEEE UPCON 2025 — International Conference on Ubiquitous Power Electronics & Communication Systems
IIT-BHU, Varanasi, India · Published & Presented · IEEE Xplore indexed
Published
2025

CryptoFusion: Temporal Graph-Aware Transformer for Cryptocurrency Forecasting

IEEE Big Data Conference
Under Review · Expected 2025
Under Review
Projects06/09

Selected Projects

01
GenAI · Autonomous Vehicles

SafeNav-RAG

ProblemLLM hallucination in safety-critical AV decisions
SolutionHybrid FAISS + BM25 RAG with FastAPI microservice
Impact~55% hallucination ↓ · <100ms · IEEE UPCON 2025
LangChainFAISSFastAPIDocker
GitHub ↗
02
Finance · Forecasting

CryptoFusion

ProblemMulti-modal, temporally correlated volatile crypto markets
SolutionTransformer + TGN + FinBERT sentiment fusion
Impact8.3% RMSE ↓ · +0.29 Sharpe ratio · IEEE Submitted
PyTorchTGNFinBERTSHAP
GitHub ↗
03
Edge AI · Security

Offline LLM Suite on RISC-V

ProblemLLM inference on edge without cloud APIs
SolutionLLaMA, Phi-4, Gemma, LLaVA with 4-bit quantization
Impact1M+ embedding retrieval · Fully offline · CPU-only
RISC-VPyTorchFAISSQuantization
GitHub ↗
04
API · Multimodal

FastAPI LLM Microservice

ProblemProduction multimodal LLM serving with low overhead
SolutionAsync FastAPI + Docker + GitHub Actions CI/CD
ImpactText + vision inference · Unit-tested · Production-ready
FastAPIDockerPhi-4Async Python
GitHub ↗
05
Vision · Retrieval

CLIP Vector Search System

ProblemOffline multimodal image search without cloud
SolutionCLIP embeddings + FAISS for offline recognition
ImpactUnit-tested pipelines · CPU-only deployment
CLIPFAISSPython
GitHub ↗
06
Weather · Deep Learning

CONUS Thunderstorm Predictor

ProblemContinental-scale thunderstorm nowcasting
SolutionCNN-LSTM on terabyte-scale NOAA/NEXRAD/GOES-16
ImpactCONUS scale · HPC/GCP pipelines · Active research
CNN-LSTMTensorFlowHPCNOAA
GitHub ↗
Career Timeline07/09

Journey

2019
Founded Softech Solutions
Started IT consultancy in Mumbai. Delivered cloud infrastructure, networking, and digital transformation for retail, education, and banking clients. 80+ POS terminals. 99.5% uptime.
2022
B.E. Computer Engineering — GPA 9.5/10
Graduated from Mumbai University with top honors. AWS, Microsoft, and Cisco certifications. Led student association as President — 40% participation increase.
Dec 2022 – Feb 2024
Cloud System Administrator — Katalyst / Apps4Rent
Microsoft Gold Partner. Automated AWS/Azure infra via Python/Bash. Prometheus/Grafana monitoring. Applied LLM-integrated security tools. Earned 100+ Google reviews.
Aug 2024
MS Computer Science — Texas A&M University–CC
Joined graduate program. Focus: AI/ML research, distributed systems, edge AI security. VP of International Student Organization.
Jan 2025
Graduate Research Assistant — AI Security Lab
CAHSI–Google research on RISC-V multi-enclave TEE. Building CNN-LSTM for CONUS thunderstorm prediction. RAG pipelines for AV decision systems.
2025
IEEE UPCON 2025 — SafeNav-RAG Published at IIT-BHU
First IEEE publication accepted and presented. CryptoFusion submitted to IEEE Big Data. 8+ open-source AI repositories released.
Next →
PhD / AI Research Engineer Role
Targeting PhD positions at top research universities and AI Research Engineer roles at Google DeepMind, OpenAI, NVIDIA Research, Meta AI, and leading labs.
Technical Writing08/09

Technical Articles

Edge AI · Security

Securing AI Models on RISC-V with Multi-Enclave TEE

A deep dive into implementing Keystone TEE on SiFive Unmatched hardware, autoencoder anomaly detection, and protecting ML inference from adversarial attacks at the edge.

✓ Draft Complete
RAG · LLMs · Production

Optimizing RAG Latency for Autonomous Systems

How we achieved sub-100ms RAG inference for AV decision-making — hybrid FAISS + BM25 retrieval, embedding pipeline optimization, and microservice architecture tradeoffs.

✓ Draft Complete
Edge AI · Deployment

Running LLMs Offline on Edge Devices (CPU-Only)

Practical guide to deploying LLaMA, Phi-4, and Gemma on resource-constrained RISC-V hardware using 4-bit quantization, FAISS, and Hugging Face transformers without cloud APIs.

Coming Soon
Deep Learning · Weather

Scaling CNN-LSTM from Regional to CONUS Thunderstorm Prediction

Expanding spatiotemporal deep learning models from 400km² South Texas to continental U.S. scale — data engineering, model architecture changes, and HPC pipeline design.

Coming Soon
Finance · ML

Temporal Graph Networks for Crypto Market Forecasting

How combining TGNs, transformers, and FinBERT sentiment creates a richer representation of crypto market dynamics than any single modality alone.

Coming Soon
LLMs · Agents

Building Multi-Agent LLM Orchestration with LangGraph

Practical patterns for agentic AI workflows — routing, memory, tool use, and failure recovery in production LangGraph-based multi-agent systems.

Coming Soon
Skills09/09

Technical Skills

LLMs & GenAI
GPT-4ClaudeGeminiLLaMAPhi-4GemmaLLaVAFinBERTPrompt EngineeringLoRA4-bit Quantization
RAG & Retrieval
LangChainLlamaIndexLangGraphFAISS (IVF-PQ)BM25Hybrid RetrievalpgvectorPineconeCLIPEmbedding Models
ML / Deep Learning
PyTorchTensorFlowHugging FaceCNN/LSTMTransformersTGNAutoencodersSHAP
Languages & APIs
PythonC/C++JavaTypeScriptBash/ShellSQLFastAPIAsync PythonNode.jsREST APIs
Cloud & DevOps
AWSGCPAzureDockerKubernetesGitHub ActionsTerraformPrometheusGrafanaNagios
Edge AI & Security
RISC-VKeystone TEESiFive UnmatchedOn-Device LLMMulti-Enclave TEEEdge InferenceFederated Learning
Databases
PostgreSQLMySQLMongoDBRedisDynamoDBCassandraVector DBs
CS Fundamentals
Distributed SystemsMicroservicesOOPDesign PatternsMulti-threadingData StructuresAlgorithmsSDLCAgile
Experience

Work History

Jan 2025
– Present
Corpus Christi, TX
Research

Graduate Research Assistant

Texas A&M University – Corpus Christi · Software & System Security Lab
  • Designed scalable distributed ETL pipelines on HPC/GCP for terabyte-scale meteorological datasets (NOAA HRRR, GOES-16 GLM, NEXRAD) with fault-tolerant high-throughput ingestion.
  • Built microservice-based LLM inference system (FastAPI + Docker) with async REST APIs; CI/CD via GitHub Actions; achieved sub-100ms latency — published IEEE UPCON 2025.
  • Expanding CNN-LSTM thunderstorm prediction from South Texas (400 km²) to full CONUS scale with multi-source data integration.
  • Implemented multi-enclave TEEs on RISC-V (SiFive Unmatched, Keystone) with autoencoder anomaly detection for secure edge AI — CAHSI–Google project.
  • Built RAG pipelines integrating FAISS vector search, embedding models, and LLM inference for AV decision-support systems.
FastAPIPyTorchLangChainFAISSDockerHPC/GCPRISC-VCNN-LSTM
Dec 2022
– Feb 2024
Navi Mumbai, India
Industry

Cloud System Administrator

Katalyst Business Solutions (Apps4Rent) · Microsoft Gold Partner, CSP
  • Automated cloud infrastructure provisioning with Python/Bash across AWS and Azure; implemented Prometheus/Grafana monitoring for production observability under strict SLAs.
  • Administered Exchange Online, Citrix XenApp/XenDesktop, Cisco RV firewalls, Windows Server 2016/2019/2022 (AD, DNS, DHCP, IIS, Group Policy).
  • Applied LLM-integrated security tools to improve enterprise workflows; earned 100+ Google reviews for technical excellence.
  • Monitored data centers with Nagios & Prometheus; automated tasks reducing incident response time significantly.
AWSAzurePrometheusGrafanaPowerShellWindows ServerCitrix
2019
– Jul 2024
Mumbai, India
Founder

Founder & Cloud Infrastructure Consultant

Softech Solutions — Mumbai, India
  • Architected fault-tolerant cloud infrastructure for retail, education, and banking — 99.5% uptime across 80+ POS terminals; WAN+VPN for HQ↔branch connectivity (Life Panacea 9 branches).
  • Designed secure server cluster with real-time sync and automated backup/DR for Jay Ma Ambe Co-Op Bank — 100% data integrity across branches.
  • Set up computer labs supporting 200+ students across AI, ML, Data Science, and Programming courses (FlyIT Infotech).
  • Deployed POS, biometric access, CCTV, sales dashboards — 18% transaction speed ↑, 40% downtime ↓ (Titan Showroom).
Windows ServerVPN/WANNode.jsReactHTML/CSS/JS
Education

Academic Background

Aug 2024 – Present

Master of Science
Computer Science

Texas A&M University – Corpus Christi
Corpus Christi, TX · Research: AI Security, LLM Systems, Spatiotemporal ML
Jun 2019 – Sep 2022

Bachelor of Engineering
Computer Engineering

G.V. Acharya Institute of Engineering & Technology
Mumbai University, India
GPA: 9.5 / 10
Professional Certifications

Certifications

AWS – Cloud Essentials for Business Leaders (May 2022)
Microsoft – Security, Compliance & Identity Fundamentals (Jul 2022)
Cisco – Networking Essentials (May 2022)
Anudip & METTL – Advanced Java Web Programming (Jun 2022)
ICE Technology Lab – Python: Basic to Advanced (Mar 2018)
Contact
Let's
BuildTogether.
Currently Available
Based in Corpus Christi, TX — Open to Relocation & Remote Worldwide.
Seeking: AI Research Engineer · ML Engineer · Research Scientist · PhD Positions
Targets: Google DeepMind · OpenAI · NVIDIA Research · Meta AI · Top University Labs