Akash Borigi

About Me

Research-Oriented AI/ML Scientist

I am a doctoral-level AI/ML researcher dedicated to advancing trustworthy, interpretable agentic AI systems through rigorous academic inquiry. My PhD-level thesis on "Architecting, Understanding, and Securing Agentic AI Systems" bridges academic rigor with pragmatic industry applications—focusing on alignment safety, LLM systems architecture, and enterprise-scale AI deployment.

📈 Research Achievements

250×217

Intent-Instruction Alignment

94.25%

Classification Accuracy

92%

Prediction R²

4

Research Publications

Key Focus Areas

🎯 Agentic AI Safety - Intent-instruction taxonomy and alignment frameworks
🤖 LLM Systems - Production-grade multi-agent architectures
📊 Applied ML - Rigorous modeling with statistical validation
🏆 Research Output - ESANN 2026 accepted, peer-reviewed publications

Technical Skills & Expertise

💻 Programming Languages

Core Languages

Python SQL R Java

🧠 AI & Machine Learning

Core ML Concepts

LLMs Transformers NLP Classification Clustering

ML Techniques

Regression Reinforcement Learning t-SNE TF-IDF

Frameworks & Libraries

TensorFlow Keras scikit-learn Pandas NumPy FAISS

🤖 LLM & Agentic AI Systems

Frameworks & Platforms

CrewAI LangChain LangGraph Google ADK

LLM Tooling

LiteLLM AWS Strands Deep Agents

Research Focus

Alignment Safety Agentic Workflows Multi-Agent Systems

☁️ Data, Cloud & DevOps

Databases

PostgreSQL MySQL MongoDB

Cloud & DevOps

AWS Docker Git

Systems & Tools

VS Code Cursor GitHub Copilot Claude Code

🔧 Systems & Collaboration Tools

Miro Confluence OPCAT CATIA

Research Experience

Aug 2024 – Present

Graduate Researcher, Agentic AI Systems

Missouri University of Science and Technology | Thesis-Focused Work

🎓 Doctoral-Level Research

Built interpretable alignment gate using 250-intent × 217-instruction taxonomy
Designed deterministic thresholding rules for safety-critical agentic workflows
ESANN 2026 Accepted publication on agentic AI guardrails
Addressed hallucination, goal drift, and unauthorized actions in multi-step workflows
Supervised thesis work with rigorous academic methodology and reproducibility

May 2025 – Aug 2025

Agentic AI Researcher

AICEBERG | Production Systems & Research

Built multi-agent enterprise IT workflow with LLMs, memory, tools, and HITL
Integrated LDAP, SQL, and ServiceNow using RAG-based retrieval with rigorous testing
Improved system transparency and traceability through reproducible workflows
Demonstrated commitment to production-ready, maintainable code
Collaborated across teams to ensure research translated to deployment

Jun 2023 – Aug 2024

Data Science Researcher

Antunes Lab, University of Houston | NIH-Funded Research

🔬 Research & Publication Contribution

Secured seed grant for peptide-based ligand benchmarking project
Reduced 50,048 → 3,976 samples through AI-driven, validated data workflows
Applied PyRosetta Monte Carlo simulation to 957 structured molecular samples
Executed 20 scoring functions in Docker for reproducible ligand-receptor evaluation
Contributed to research publications in computational drug discovery

Apr 2023 – May 2023

AI Research Assistant

University of Houston | Research Infrastructure

Built AI-assisted publication data web program with Python backend
Consolidated metadata from 86 journals and 180 conferences with meticulous data handling
Enabled streamlined research access for 3+ research groups
Demonstrated full-stack commitment: backend, frontend, deployment, and maintenance

Research & Collaborative Projects

Each project reflects rigorous methodology, reproducibility, and commitment to advancing AI/ML research

🔬

Comprehensive Deep Learning Analysis for Prompt Intervention and Toxicity Assessment

Aug 2024 – Dec 2024

Designed and evaluated CNN, RNN-LSTM, GRU, and TCN models with GloVe embeddings and rigorous NLP preprocessing
Built ensemble voting classifier achieving 94.25% test accuracy and 0.206 test loss
Implemented reproducible evaluation methodology with systematic cross-validation and performance metrics

Deep Learning NLP Ensemble Methods

🤖

Design and Assessment of an AI Assistance Agent System

Aug 2024 – Dec 2024

Designed AI assistance agent architecture for query processing, context management, decision-making, and action execution using OPCAT
Built fuzzy-logic assessment framework to rigorously evaluate scalability, reliability, responsiveness, adaptability, and security
Developed systematic methodology for agentic AI system evaluation and validation

Agentic AI Systems Design Assessment Framework

🏗️

ChatGPT System Modeling and Architecture Analysis

Jan 2025 – May 2025

Modeled ChatGPT-like system in CATIA using SysML to capture requirements, architecture, interfaces, and behavioral specifications
Built comprehensive SysML diagrams: BDD, IBD, activity, and state diagrams for core AI subsystems, APIs, security, monitoring, and user interaction
Developed rigorous systems engineering approach to AI application architecture and design documentation

Systems Modeling SysML Architecture

📊

Pioneering the Drill using Comparative Predictions (NSF Apprenticeship)

May 2023 – May 2024

Applied 2 normalization and 3 transformation techniques to improve model performance by 17% across multiple regressors
Achieved 92% R² using Random Forest on 7,310 drilling records with rigorous feature engineering and randomized cross-validation up to 10 splits
Demonstrated reproducible predictive modeling methodology on real industrial data with comprehensive validation

Predictive Modeling Feature Engineering NSF Funded

Publications & Research Output

Peer-reviewed research demonstrating rigorous methodology and scientific rigor

✅ Accepted

The Alignment Gate: Intent and Instruction Guardrails for Agentic AI

ESANN 2026 | Borigi, A., Lindner, P., Schlager, A., Shoaib, S., Lekkala, R., Bhamidipati, S. S., and Lendasse, A.

Presents rigorous interpretable alignment framework using empirically-validated 250×217 intent-instruction taxonomy for safer agentic AI execution. Thesis-quality research combining theory with practical agent traces and safety validation.

Under Review

Hierarchical Channel Stacking: A Structured Decision Framework for AI-Generated Image Detection

ICANN | Shoaib, S., Borigi, A., Schlager, A., Lekkala, R., Bhamidipati, S. S., Lendasse, A. and Lindner, P.

Contributes to structured, interpretable approaches for detecting synthetic media and AI-generated content.

Under Review

A New Methodology to Prevent Text-to-SQL Misalignment

ICANN | Lekkala, R., Lindner, P., Schlager, A., Shoaib, Borigi, A., S., Bhamidipati, S. S., and Lendasse, A.

Rigorous research focusing on reducing hallucination and misalignment risk in text-to-SQL generation for enterprise queries. Demonstrates commitment to solving real production challenges through peer-reviewed research.

In Preparation

Multi-Instruction Alignment for Safe and Trustworthy Agentic AI

ASEM 2026 | Borigi, A., Schlager, A., Lindner, P., Lendasse, A.

Extends alignment research to multi-step planning responses with rigorous safety analysis. Advances theoretical and practical understanding of agentic AI safety through peer-reviewed research.

Published

Machine Learning Algorithms on Public Mental Health

IJSE 2022 | Rao, K. R., Borigi, A., Annam, V., Devalla, S., and Sree Chandana, M.

Introduces rigorous framework for mental health professionals to tackle challenges using data science methodologies. Multi-author collaborative research demonstrating commitment to real-world impact.

Education & Academic Background

🎓 M.S. Systems Engineering (Thesis)

Missouri University of Science and Technology

Expected Aug 2026 • Rolla, MO

PhD-level thesis research: "Architecting, Understanding, and Securing Agentic AI Systems"

Doctoral-level research on interpretable alignment frameworks, safety-critical agentic workflows, and trustworthy multi-agent systems. Full thesis focus with peer-reviewed publication output (ESANN 2026 accepted).

🎓 M.S. Engineering Data Science

University of Houston

Aug 2024 • Houston, TX

Advanced coursework in ML, data science, data mining, and analytics. Completed 30 graduate credits, supported by a Merit Dean Scholarship. Collaborated with the Antunes Lab on an NIH grant–related drug discovery project involving AI-driven peptide and ligand analysis.

🎓 B.Tech. Electronics & Computer Engineering

Koneru Lakshmaiah Education Foundation

Jun 2022 • Vijayawada, India

Explored a blend of computer science coursework, including AI, data science, and web applications, alongside core electronics subjects such as digital signal processing and analog circuit design. This interdisciplinary foundation culminated in a capstone project that led to journal-recognized research work.

Let's Collaborate

I'm actively seeking opportunities to contribute my research expertise and collaborative commitment to leading AI/ML teams. Whether you're advancing agentic AI safety research, building production LLM systems, or solving critical applied ML challenges—I'm ready to make a dedicated impact.