Professional Odyssey & Experience

A timeline of engineering complex AI systems and architectural innovation.

JRat’s Studio

Software Engineer, FreelanceRemote

ActiveMarch 2026 - Present
AWS (Batch, ECS, RDS, Cognito, S3)Next.jsPythonPolarsGo
  • Building a High-Throughput IPO Allotment & Reconciliation Engine: Developed a Python-powered engine to automate complex lottery and proportionate allocation logic for Mainboard and SME IPOs, achieving a benchmark of 20 million records in ≤90 minutes to meet SEBI’s T+3 mandate.
  • Engineering a Secure External Integration Layer: Connected with 8 national financial systems (BSE, NSE, NSDL, CDSL) via SFTP and REST APIs, implementing the mandatory UDIFF protocol and SEBI CSCRF cybersecurity standards.
  • Architected and Built an ERP System: Developed a custom ERP system to streamline internal operations, financial tracking, and resource management for the studio.
Omara Technologies

Founding Full-Stack AI EngineerIndia

October 2024 - January 2026
GoPythonNext.jsPostgreSQLNeo4jAWS (EKS, Cognito, S3)DockerLLMs (Gemini)GraphRAGLangChain/LangGraph
  • Architected and Deployed a Cloud-Native PDF Annotation Platform: Built a scalable, enterprise-grade system using Go (Golang) and PostgreSQL, leveraging AWS Cognito for secure, role-based access control (RBAC) and AWS S3 for persistent document storage
  • Developed a High-Concurrency Task Engine: Managed 100+ RESTful endpoints with a task distribution system supporting parallel and series strategies, utilizing Docker for containerization and optimized for deployment on AWS ECS/EKS with CloudWatch monitoring
  • Engineered a Non-Deterministic Bipartite Instance Alignment & 4-Tier Consensus Framework: Resolved inter-annotator discrepancies by formulating a global optimization problem using the Hungarian Algorithm (Kuhn-Munkres); implemented order-independent matching for multi-instance entities by minimizing a cost matrix derived from Value Similarity (60%), Spatial Proximity/IoU (30%), and Labeler Agreement (10%)
  • Built Knowledge Graph Platform (Docunexus): Orchestrated an end-to-end GraphRAG pipeline using Google Gemini 1.5 Pro for structured extraction and Neo4j for entity-relationship mapping, enabling deep discovery across unstructured legal document collections
  • Architected an Advanced Hybrid Search & GraphRAG System: Implemented a 7-node LangGraph workflow that executes parallel semantic vector searches and NL-to-Cypher conversions, providing conversation-aware responses with automated source citations
  • Optimized AI Infrastructure & Costs: Reduced LLM API operational costs by 70% by engineering an Intelligent Context Caching system for Gemini that prioritized system instructions and managed token limits for batch processing
Independent Tutoring

DSA & Machine Learning InstructorIndia

January 2024 - September 2024
Data StructuresAlgorithmsMachine LearningPythonProblem SolvingMentoring
  • Mentored and taught Data Structures & Algorithms (DSA) and Machine Learning concepts to diverse students, ranging from beginners to advanced learners preparing for technical interviews
  • Designed and delivered personalized curriculum tailored to individual student needs, covering topics from fundamental data structures (arrays, trees, graphs) to advanced algorithmic paradigms (dynamic programming, greedy algorithms)
  • Guided students through practical ML projects including supervised learning, neural networks, and model optimization, helping them build real-world applications and portfolios
  • Achieved high student success rate with multiple students securing placements at top tech companies after intensive coaching sessions
  • Developed custom problem sets and practice materials focusing on competitive programming platforms like LeetCode and Codeforces, significantly improving students' problem-solving speed and accuracy
GeekforGeeks

Remote Content WriterIndia

October 2022 - March 2024
Technical WritingData StructuresAlgorithmsComputer VisionData Science
  • Contributed DSA problems and editorials for the GeekforGeeks platform, helping thousands of learners understand complex algorithmic concepts
  • Wrote technical blogs on various cutting-edge technologies including computer vision, data science, and advanced algorithms
  • Created educational content that simplified complex technical topics, making them accessible to learners at different skill levels
HackerEarth

Remote Problem SetterIndia

April 2021 - August 2021
Algorithm DesignProblem SettingCompetitive ProgrammingTest Case Generation
  • Designed and created original algorithmic problems for HackerEarth's problem library, used by thousands of developers for practice and skill assessment
  • Developed comprehensive test cases and edge cases to ensure problem robustness, covering various difficulty levels from easy to advanced
  • Crafted problems specifically tailored for company recruitment drives and coding contests, helping companies evaluate candidates' problem-solving abilities effectively
  • Collaborated with the editorial team to write detailed editorials and solutions explaining optimal approaches, time complexity analysis, and implementation strategies
  • Contributed to the platform's growing repository of high-quality competitive programming challenges, enhancing the learning experience for the developer community

Core Competencies

A holistic skillset bridging advanced AI research with robust software engineering.

AI & ML

LLMs, Computer Vision, ML Pipelines

Cloud Native

AWS, Lambda, Docker, Microservices

Full-Stack

Next.js, Go, Python, TypeScript

Data Systems

PostgreSQL, MongoDB, Neo4j