Curriculum Vitae
| Resume |
PhD candidate and AI safety researcher with 10+ years of experience, specializing in physics-informed deep RL, uncertainty quantification, and adversarial robustness for critical infrastructure, committed to building AI that is safe, auditable, and deployable in the real world.
Professional Experience
Algoverse AI Research
| AI Safety Research Fellow | Jan 2026 — Present |
- Conducting agentic AI safety research across evaluations, adversarial robustness, and scalable oversight.
- Designing behavioral and mechanistic evaluations to detect deceptive alignment, reward hacking, and obfuscated internal states in LLM agents.
- Collaborating with researchers and mentors to design safety benchmarks and diagnostic experiments.
BlueDot Impact
| AI Safety Fundamentals — Course Participant | Jan 2026 — Feb 2026 |
- Completed BlueDot Impact’s 6-week AI Safety Fundamentals cohort, covering core technical and governance dimensions of AI safety including alignment, interpretability, robustness, and oversight.
- Studied key alignment paradigms — RLHF, Constitutional AI, scalable oversight, and debate — alongside mechanistic interpretability techniques and evaluation frameworks for frontier models.
- Engaged in structured readings, facilitated discussions, and collaborative problem-solving with a global cohort of researchers and practitioners working on safe and beneficial AI development.
National Renewable Energy Laboratory (NREL)
| Machine Learning Engineer (Intern) | May 2024 — Jan 2025 |
- Developed novel machine learning models for automated network topology inference and resilient control policy optimization for complex distributed systems under extreme and uncertain operating scenarios.
- Designed and implemented Bayesian neural network based semi-supervised learning frameworks with explicit uncertainty estimation to learn from limited and unreliable labeled data in energy distribution networks, achieving up to 98% improvement in model accuracy across varying label-availability and data-quality settings.
- Co-authored the paper “Advanced Semi-Supervised Learning with Uncertainty Estimation for Phase Identification in Distribution Systems,” accepted at IEEE Power & Energy Society General Meeting (PES GM) 2025, demonstrating robust phase identification under noisy measurements and scarce ground-truth labels.
- Deployed the trained ML models into a practical workflow for distribution utilities, including data preprocessing pipelines, model serving, and evaluation; implemented containerized deployment and API-based integration with existing grid analytics tools to enable reproducible, scalable, and operator-ready decision support.
Comcast
| Software Engineer | Jul 2019 — Feb 2020 |
- Built and scaled real-time data pipelines using Amazon Kinesis, RabbitMQ, and microservices, processing 1TB+ streaming data/day to support fraud detection and system monitoring at production scale.
- Developed and deployed ML-based anomaly and behavior-analysis models, incorporating temporal features and probabilistic scoring, leading to a 70% reduction in fraudulent activity through early-signal detection.
- Delivered low-latency analytics dashboards with PrestoDB, Athena, and Python, enabling real-time visibility into network performance and fraud patterns and improving cross-team decision speed and operational response.
- Collaborated with cross-functional teams (security, data engineering, product) to integrate fraud-intelligence signals into production systems, improving detection latency, system resilience, and customer-impact mitigation.
IBM
| Software Engineer | Jan 2019 — Jun 2019 |
- Optimized large-scale cloud infrastructure on OpenShift, implementing adaptive auto-scaling and resource-allocation policies that reduced operational costs by 30% while improving system stability.
- Built a real-time monitoring and observability platform using Grafana, Flask, and distributed exporters, providing actionable visibility across 100+ cloud servers and critical microservices.
- Designed and automated performance-monitoring and alerting pipelines, cutting incident response time by 60% through intelligent thresholds, anomaly alerts, and integrated on-call workflows.
Hewlett Packard Enterprise (HPE)
| Software Engineer | Apr 2017 — Dec 2018 |
- Led the zero-downtime migration of critical enterprise applications from HPI → HPE domains, coordinating infrastructure, DNS, and service-cutover workflows to ensure seamless continuity for all users.
- Built and integrated OAuth 2.0–based authentication and secure REST APIs using Spring Boot, hardening access controls and improving reliability for applications serving 50K+ active users.
- Designed and deployed a microservices architecture across Apache and WebLogic servers, optimizing service boundaries and request flows to achieve a 40% improvement in system response time.
Tata Consultancy Services (TCS)
| System Engineer | Jul 2012 — Dec 2015 |
- Engineered high-throughput ETL pipelines for large-scale data-warehouse integration, reliably processing 100GB+ of data per day and improving downstream analytics latency.
- Optimized database performance through advanced SQL tuning, indexing, and query-plan diagnostics, reducing execution times by 70% across critical workloads.
- Delivered $100K in annual cost savings by leading database-optimization and storage-efficiency initiatives, earning an organizational Excellence Award for impact on operational efficiency.
Education
Iowa State University
| Ph.D. in Computer Science (Minor: Statistics) | 2020 — 2026 (Expected) |
- Research Focus: Deep Reinforcement Learning, Physics-Informed AI, Uncertainty Quantification, Bayesian Modeling, Secure & Robust Learning, and LLM-Driven Autonomous Agents. Work spans critical-infrastructure optimization, safety-critical control, and large-scale distributed systems.
- Technical Contributions: Developed physics-informed DRL algorithms for real-time control, Bayesian neural networks with uncertainty estimation for limited/unreliable data, adversarial robustness frameworks for cyber-physical systems, and LLM-augmented decision-making for energy and autonomous systems.
- Graduate Coursework:
- Deep Learning, Natural Language Processing, Advanced Machine Learning, Computer Vision, AI for cybersecurity, Algorithms, Database Systems, Computer Networking
- Statistical Theory, Empirical Methods, Experimental Design, Data Analysis and Visualization
Teaching Experience
Iowa State University
| Teaching Assistant | 2020 — 2026 |
Department of Computer Science
- Supported multiple undergraduate and graduate courses, including Software Development Practices, Database Systems, and Spreadsheets—impacting 300+ students across semesters.
- Led weekly labs and office hours, guiding students through debugging, system design reasoning, data modeling, and code quality challenges; strengthened problem-solving skills across diverse cohorts.
- Designed programming assignments, quizzes, and real-world case studies aligned with industry workflows, agile practices, and modern software engineering tools (Git, CI/CD, databases).
- Mentored students on semester-long capstone projects, coaching teams on architecture decisions, sprint planning, testing, and documentation to simulate professional engineering environments.
Research Experience
Iowa State University
| Research Assistant | Aug 2022 — Jul 2025 |
Physics-Informed Deep Reinforcement Learning for Critical Infrastructure Systems
- Conducted research on physics-informed deep reinforcement learning (DRL) for large-scale distributed networks, advancing intelligent resource management and security for critical infrastructure
- Applied computational DRL algorithms in smart energy systems, optimizing real-time control policies to minimize voltage violations, reduce power loss, and improve system stability across diverse operating conditions.
- Developed physics-informed actor–critic algorithms that embed domain constraints into the learning process, achieving 30% higher resource-allocation efficiency and significantly reducing violations in complex networks.
- Designed adversarial attack detection and mitigation frameworks for grid-scale AI models, performing systematic stress testing and implementing defensive DRL techniques to enhance robustness against security threats.
- Created transfer-learning pipelines enabling DRL agents to generalize across networks of different sizes and topologies, cutting retraining time by 40% when deploying to new environments.
- Built a Python-based simulation and real-time control framework integrating OPAL-RT hardware with OpenDSS and distributed system components to support HIL (hardware-in-the-loop) experiments.
- Integrated LLM-driven reasoning to support adaptive control, predictive optimization, and human-AI collaboration inside simulation environments, improving interpretability and situational awareness.
| Research Assistant | Aug 2020 — Jul 2022 |
Deep Reinforcement Learning & Safety-Critical AI for Autonomous Systems
- Conducted research on deep reinforcement learning and safety-critical autonomy, focusing on robust perception, control, and sequential decision-making in high-stakes environments.
- Utilized CARLA simulator to develop end-to-end autonomous driving stacks, including vision-based perception, object detection, trajectory planning, and policy learning under complex traffic dynamics.
- Applied deep computer vision models for object recognition, semantic segmentation, and multi-sensor fusion, enabling reliable situational awareness and improving downstream control performance in autonomous driving systems.
Skills
Programming
Python, R, Java, C++, SAS, MATLAB, SQL, HTML/CSS, JavaScript (Node.js, React)
Machine / Deep Learning
scikit-learn, TensorFlow, PyTorch, pandas, NumPy, Matplotlib, Seaborn, Gym, RLlib, Stable-Baselines3
LLMs & NLP
OpenAI API, Hugging Face Transformers, LangChain, LlamaIndex, RAG (retrievers, chunking, reranking), vector DBs (FAISS, Chroma, Pinecone), prompt engineering & structured outputs, evaluation (Ragas, Promptfoo)
Agentic Systems & Orchestration
LangGraph (stateful workflows), LangChain Agents (ReAct/MRKL/tools), function/tool calling, multi-agent design, planning & memory, tool-use (search/code/execution), guards & grounding
MLOps & Deployment
MLflow, Weights & Biases, Docker, Kubernetes, FastAPI, gRPC, CI/CD (GitHub Actions, Jenkins, CircleCI), model versioning, scalable inference, monitoring & drift detection
AI Safety & Robustness
Safety evaluations & red-teaming pipelines, gradual rollout strategies, alignment & policy-enforcement guardrails, interpretability tooling (SHAP, LIME, Captum), adversarial robustness frameworks (Foolbox, RobustML), anomaly & drift monitoring, reliability & stress testing for safety-critical deployments.
HPC & Big Data
Hadoop, Hive, Spark, Kafka, Kinesis, Presto, Athena, distributed computing: SLURM, MPI, OpenMP
Simulation & Modeling
OPAL-RT (HIL), OpenDSS (power systems), CARLA (autonomous driving), CityLearn, cyber-physical system simulation & real-time control frameworks
Optimization
Gurobi, Pyomo, BoTorch, Optuna, Hyperopt
Visualization & GIS
Tableau, ArcGIS, Leaflet, Plotly, Dash
Cloud & DevOps
AWS (EC2, S3, Lambda, EKS), GCP, DigitalOcean, Terraform, Docker, Kubernetes, Git, Jenkins, CircleCI
Honors & Awards
- Selected, Seventh Workshop on Autonomous Energy Systems @ NREL (2024)
- Selected, ByteBoost Workshop on Accelerating HPC Research Skills (2024)
- Selected, Oxford Machine Learning Summer School (OxML) (2022)
- Excellence Award, Database Optimization @ TCS
- 2nd Place, BAJA SAE India (Safest Terrain Vehicle Category, National Level)
- Won multiple robotics competitions at inter-university technical festivals.
Service
Reviewer:
- IEEE Transactions on Industrial Informatics (2025)
- Conference on Neural Information Processing Systems (Ethics)(2025)
- IEEE Transactions on Neural Networks and Learning Systems (2024)
- IEEE PES GM, Grid Edge & ISGT (2023, 2024)
Mock Interviewer: Supporting underrepresented minorities in tech.
Volunteer, Prayaas India (BIT): NGO providing quality education to underprivileged children in slums and villages.
