Overview
Ashwath Raj Mulakalapalli is a Computer Science Engineering student at SRM University–AP working across Applied AI, intelligent software, quantum computing, machine learning, distributed systems, optimization, and core Computer Science.
His work is centered on understanding how complete systems behave under real constraints. Rather than focusing only on individual models or features, he is interested in how data, software, algorithms, infrastructure, and people fit together to produce reliable results.
Public profile: this page is a concise technical overview. Detailed professional work and project updates are available through LinkedIn. Source code and selected implementations are available through GitHub.
“The focus is not simply on building features; it is on understanding the constraints, decisions, and architecture that make systems work.”
Education
| Institution | Programme | Period | Academic record |
|---|---|---|---|
| SRM University–AP | B.Tech, Computer Science Engineering | 2025–present | 9.69 / 10 current CGPA |
| SRM University–AP | Semester 1 | Completed | 9.78 CGPA |
| SRM University–AP | Semester 2 | Completed | 9.61 SGPA |
Academic work has been pursued alongside research, internships, hackathons, independent projects, algorithmic practice, and continued study of quantum computing.
Experience
RURO — Program Development Intern
Worked on the GOPIKA AI initiative, contributing to systems that brought together CCTV, GPS, farm data, and operational information. The experience involved understanding how AI systems have to work with different types of real-world data and practical implementation constraints.
Broader work around GOPIKA included local AI, document processing, information retrieval, streaming, multimodal interaction, and system integration.
- Architected and developed the first functional GOPIKA prototype for AI-powered agricultural decision support.
- Built multimodal AI pipelines integrating RAG, LLM inference, OCR, avatar rendering, subtitle generation, and execution monitoring.
- Engineered deterministic retrieval, observability, and validation systems with context grounding, fallback strategies, runtime diagnostics, and adversarial testing.
- Contributed to architecture reviews, deployment planning, stakeholder presentations, and multimodal integrations across CCTV, GPS, livestock, and operational datasets.
GYTWorkz — AI Engineering Intern
Worked on AI systems for enterprise use, exploring AI agents, self-improving systems, evaluation, governance, workflow automation, knowledge management, and large-language-model applications.
Work included combining several ways of working with enterprise information, improving how AI systems receive context, coordinating multiple AI components, and exploring multimodal and distributed workflows.
- Developed self-evolving AI agent architectures capable of continuously expanding organizational knowledge and improving decision-making workflows.
- Built an Agent Harness framework for orchestrating, evaluating, and monitoring autonomous agent behavior across enterprise use cases.
- Designed intelligent workforce-assistance systems enabling optimized task allocation through organizational knowledge modeling and agent-driven reasoning.
- Implemented knowledge-ingestion and retrieval pipelines allowing agents to learn company structure, staff capabilities, workflows, and operational context.
The experience shifted the focus from building isolated AI features toward designing complete systems that can be evaluated, controlled, and used in practical workflows.
Detailed professional profile →Research
Machine Learning Driven Predictive Modeling and Optimization in Biogas Systems
Co-authored and presented a research paper on machine-learning-based prediction and optimization in biogas systems.
- Developed ANN, Random Forest, SVM, and XGBoost models for predictive biogas system analysis.
- Applied GA, NSGA-II, and Pareto-front optimization for multi-objective decision modeling and yield trade-off evaluation.
- Presented findings on model optimization and validation using noisy real-world datasets at ICML-ALGO.
The work highlighted an important distinction: predicting an outcome is different from deciding what action is best when several objectives compete.
Artificial Intelligence
AI work has progressed from using individual models toward understanding the systems around them: what information a model receives, how that information is selected, how different components work together, and how results are checked.
AI Agents & Intelligent Workflows
Explored AI agents, self-improving systems, automated workflows, evaluation, human oversight, governance, organizational memory, and controlled policy changes.
Built a proof-of-concept for a controlled engineering workflow in which an AI system could help break down requirements, plan work, execute tasks, collect results, evaluate performance, and learn from previous outcomes while keeping human approval in the loop.
The main interest is not making an AI system “autonomous” for its own sake, but making it useful, traceable, measurable, and safe enough to operate within defined boundaries.
Enterprise AI & Knowledge Systems
Explored several approaches to helping AI work with enterprise information:
- RAG
- Retrieving relevant information before generating an answer.
- CAG
- Giving the model carefully prepared context.
- GraphRAG
- Using relationships between information to improve reasoning.
- Text-to-SQL
- Allowing natural-language questions over structured databases.
These approaches were studied alongside prompt design, context management, structured data, knowledge grounding, validation, and AI orchestration. A recurring concern is choosing the right approach for the question rather than forcing every problem through the same retrieval method.
Document & Content Intelligence
Worked with AI content pipelines involving document and image ingestion, OCR, preprocessing, structured outputs, and validation. The aim is to turn messy inputs into information that downstream AI systems can actually use.
Multimodal Streaming
Explored systems that stream text and audio together, including queues, backpressure, audio generation, subtitles, and time synchronization. The challenge is keeping several independently produced signals aligned while information is still being generated.
Quantum Computing
Quantum computing is a major parallel area of interest, covering quantum communication, optimization, simulation, and distributed execution.
Q-AERO
Developed a prototype distributed quantum workload system aimed at aviation optimization. The system explored splitting quantum workloads across three devices, sharing available computing resources, running Qiskit simulations, and combining the results.
The prototype was designed around the practical limitations of classical quantum simulation and investigated how distributed resources could be combined to execute workloads that become increasingly expensive as circuit size grows.
QAOA & Optimization
Worked with QAOA for optimization problems related to aviation, where factors such as fuel, time, congestion, weather, and delay can compete with one another.
BB84 & QKD
Worked with BB84 and quantum key distribution concepts, including quantum measurement, basis selection, and detecting interference through error rates.
Quantum Projects
Explored quantum random number generation, Superdense Coding, Qiskit simulation, quantum sensing, encryption, and quantum optimization.
One practical challenge in quantum simulation is memory. A classical computer must represent a quantum state using a state vector whose size grows exponentially with the number of qubits. This makes distributed execution and efficient resource use important when working beyond small simulations.
Systems Engineering
Core systems interests include Linux and Ubuntu, networking, APIs, databases, concurrent processing, distributed execution, resource management, and software architecture.
This systems perspective is also applied to AI. A model is treated as one part of a larger pipeline involving data, context, storage, processing, execution, validation, and monitoring.
Projects
GOPIKA — Multimodal AI Assistant
Worked on an AI assistant ecosystem involving local language models, information retrieval, document processing, streaming, real-time interaction, and multimodal workflows.
Q-AERO — Distributed Quantum Workload System
Prototype for distributing quantum workloads across multiple devices, with a focus on aviation-related optimization and practical limits of quantum simulation.
AI-Native Content Pipeline
Explored a pipeline for processing documents, images, and other content and converting them into structured information suitable for AI systems.
Water Quality Intelligence Dashboard
Re-engineered a water-quality dashboard combining geographic information with Water Quality Index modelling.
Legal Document Demystifier
Built an AI-based system for extracting information from legal documents and presenting complex material in a more accessible form.
Quantum Communication & Randomness Projects
Explored quantum communication and randomness through QKD, Superdense Coding, and a Quantum Random Number Generator prototype.
Competitions & Hackathons
Participated in 16+ national-level hackathons across AI, quantum computing, systems engineering, cybersecurity, optimization, product design, and analytical problem solving.
| Competition / Environment | Outcome / Focus |
|---|---|
| IIIT Nagpur | Finalist with a solo QRNG-oriented prototype |
| Amaravati Quantum Valley | Semifinalist; quantum systems and application-focused work |
| My1stHack | Best Design Award; Team Lead |
| CodeFest’26 — CTF | Reverse engineering, web security, cryptography, and forensics |
| Analytical Challenge | 1st Runner-Up; reasoning and problem decomposition |
| NationBuilding Case Study Competition | Multi-stage case analysis and structured solution design |
| Quantum competitions | Quantum algorithms, communication, optimization, and system design |
Entrepreneurship
Entrepreneurship has provided a different engineering environment: communicating an idea clearly, defending technical choices, understanding feasibility, and responding to structured evaluation.
MOTIX
Pitched at state level.
MONU
Pitched to an international jury.
Experience with product framing, feasibility, presentation, and competitive evaluation.
Algorithms & Core Computer Science
Algorithmic practice is maintained through competitive programming and regular problem solving, with 350+ LeetCode problems solved.
| Area | Focus |
|---|---|
| Data structures | Arrays, strings, linked lists, stacks, queues, trees, graphs, tries, and hashing |
| Algorithms | Binary search, sliding window, two pointers, greedy methods, graph algorithms, recursion, and dynamic programming |
| Core systems | Pointers, memory management, complexity, concurrency, Linux, and networking |
| Problem solving | 350+ LeetCode problems |
The purpose of this practice is not simply the problem count. It is to build the ability to break problems down, reason about complexity, and make disciplined decisions under constraints.
Technical Profile
| Category | Technologies & Tools |
|---|---|
| Languages | C, C++, Python, Java, C#, SQL, JavaScript, HTML, CSS |
| AI & LLM Systems | RAG, CAG, GraphRAG, Text-to-SQL, AI Agents, Multi-Agent Systems, Prompt Engineering, Context Engineering, LLM Evaluation, NLP, OCR |
| AI Tools | Ollama, LangChain, CrewAI, Qdrant |
| Backend & Data | FastAPI, REST APIs, PostgreSQL, MySQL, SQLite, Neo4j |
| Systems & Infrastructure | Linux / Ubuntu, Docker, Git, GitHub, networking, distributed execution, concurrent processing |
| Frontend | React, Vite, Tailwind CSS |
| Quantum & Optimization | Qiskit, QAOA, BB84, QKD, QRNG, Superdense Coding, Genetic Algorithms, NSGA-II |
The profile is intentionally broad: programming fundamentals and algorithms form the base, while AI, systems, quantum computing, research, and product work provide different environments in which those fundamentals are applied.
Current Direction
Current interests are concentrated around Applied AI and intelligent systems, with deeper work in AI agents, enterprise knowledge systems, distributed computing, quantum computing, optimization, and production-oriented software engineering.
The direction is toward building complete systems rather than isolated demonstrations: systems that have reliable information, clear workflows, measurable results, sensible resource use, and appropriate human oversight.
The broader goal is to keep strengthening the foundations of Computer Science while exploring where AI and quantum computing can solve problems that conventional approaches struggle to handle efficiently.
External Links
- LinkedIn — professional profile and detailed public work
- GitHub — projects and source code
- LeetCode — algorithmic practice
- Résumé — PDF version of the professional profile
This page is a personal technical profile and overview. It is intentionally maintained at a public level where professional work may involve non-public implementation details.
Last structured update: August 2026.