Headline

Software Engineer @ Blue Water | CSE @ UNSW | International Scientia Scholar | Building Maritime Compliance, Self Inspection & KPI Platforms | Agentic AI • RAG • Spring Boot • Django • JavaScript • Tailwind CSS

About

Hello there! I’m Shubham Johar, a software engineer who thrives on turning ambitious ideas into practical, real‑world solutions. From improving safety at sea to advancing ethical AI, I combine deep technical expertise with a passion for solving complex problems. I’m motivated by one question: How can technology make complex problems feel simple and humane? I spearheaded the development of SIRE 2.0 SWITCH, an offline-first inspection platform revolutionizing maritime safety. Designed for tanker operations, it works offline in remote conditions (even on mobile) and dynamically tailors risk-based questionnaires to each vessel and crew rank, focusing on critical areas. The platform centralizes all documentation and uses AI (OpenCV + EasyOCR) to validate inspection photos against OCIMF standards. To make inspections smarter and more reliable, I integrated agentic AI into SIRE 2.0. The system employs an LLM-driven assistant (using LlamaIndex/LangChain) to suggest relevant questions via a chat interface and track SMS compliance. It also integrates with other systems via APIs for real-time data syncing and sends alerts for key events. Additionally, it tracks observations with full traceability to closure and provides master dashboards for KPI monitoring. Inspectors can train with built-in quizzes, and even transcribe and summarize crew interviews through OpenAI Whisper—ensuring no detail is missed. I’ve researched Kantian Automated ethics at UNSW using Dyadic Deontic Logic(more dets. below) to inform how AI should reason, plan, and act which served as the basis for my current work. Furthermore, I co-developed BOSS-GPT now utilized as Aanya - a core frontline Agentic AI System for BOSS trained on company specific data using RAG pipelines, MCP and Agentic Syncronization of ReACT Agents. It is an LLM-based co-pilot that uses advanced prompting to automate complex support queries and speed up resolutions. I also co-engineered VPS-BOSS,a fleet monitoring platform combining real-time AIS tracking with automated reporting and analytics to help compliance and fuel saving. Furthermore, i also worked on a project called Minex VR with NASA at UNSW concocting a digital twin simulation for off-earth mining operations that blended VR with engineering to explore future space resource utilization. I strongly believe in collaboration, so I welcome connections whether you’re looking to collaborate, hire, or just exchange ideas—especially in responsible AI and intelligent automation. Feel free to reach out and let’s explore how we can shape a smarter, ethical future together.

Location

Dehradun

AI stack

claude code

cursor

opencode

Work experience

Software Engineer at BOSS Voyage Optimization, Mar 2024 - Present · 2 yrs 6 mos, Dehradun, Spearheaded the design, development, and launch of an offline-first AI-driven maritime inspection platform (SIRE 2.0 SWITCH), acting as both technical lead and project manager. Led a team of 3 engineers and owned the full development lifecycle from architecture and database design to deployment, monitoring, and security. • Implemented SIRE 2.0’s risk-based inspection framework with dynamic questionnaires tailored by vessel type and crew rank, and observation traceability (repeat/shared/open/closed findings with automated reporting). • Integrated AI features using OpenCV, EasyOCR for automated photo validation, and LLM-driven agents (LangChain, LlamaIndex) for SMS compliance checks and documents chat. • Built an offline-capable self-inspection module with AES encryption, robust data recovery, and seamless sync online. • Delivered a rich feature set including OCIMF compliant PIQ and HVPQ (validated with self insp.), certificate & crew matrix management, Port State Control (PSC) reporting, training quizzes, real-time alerts, sign off, crew management and one-click report generation. • Cut inspection processing time by 6 hours per vessel and boosted inspection readiness by 56%. Improved analytics load times from 40s to <1s, saved 100+ hours for clients with an AI-driven SMS-SIRE gap analysis tool and optimized server assets by 40% by switching image storage from PNG to JPEG. • Led product demos and user onboarding, deploying the platform to 94 vessels (2,500+ users) in 6 months and driving rapid adoption through training and feedback. • Also, assisted with an internal LLM-based AI co-pilot (“BOSS-GPT”, now “Aanya”) for knowledge retrieval , decision support and “VPS-BOSS”, a project for real-time AIS tracking and smart reporting to enhance fleet compliance and visibility. •Contributed to R&D of an AI-enabled bucket-based calibration methodology, fusing sea-trial and operational data to improve prediction accuracy by 15% and achieve additional 10% fuel savings.

Student Mentor at AlgoZenith, Sep 2023 - Feb 2024 · 6 mos, Dehradun, Uttarakhand, India, • Successfully completed a rigorous and structured curriculum that included in-depth modules on Data Structures and Algorithms (DSA), Backend development (using frameworks such as Django and Node.js), Frontend engineering (focusing on React, JavaScript, and UI design principles), and System Design—covering both Low-Level Design (object-oriented design patterns, class diagrams) and High-Level Design (distributed systems, scalability, fault tolerance). • Gained comprehensive insight into industry-standard coding practices, time-space complexity optimization, database schema design, and architectural patterns used in real-world production systems. • Understood and internalized a wide range of coding paradigms and algorithmic strategies commonly tested in top-tier technical interviews, including recursion, DP, greedy, backtracking, graph theory, trees, tries, and sliding window techniques. • Actively mentored a cohort of students by offering structured guidance, one-on-one doubt resolution, and hands-on feedback for DSA problems, project architecture, and code quality improvements. • Helped students debug and optimize their solutions. • Contributed to the overall learning ecosystem by curating additional resources, explaining edge-case scenarios. • Played a key role in fostering a learning-first environment by encouraging logical thinking, clean coding practices, and efficient problem-solving strategies tailored to competitive coding and technical hiring challenges.

Health and well-being at Career Break, May 2023 - Aug 2023 · 4 mos, Dehradun, Uttarakhand, I dedicated this period to comprehensive health management following prolonged challenges with multiple health issues and post-surgery recovery. This time allowed me to fully regain my strength and focus on establishing a foundation for sustained long-term well-being.

Research Internship | Possibility of Concocting a Kantian Automated Machine at UNSW, Sep 2022 - Dec 2022 · 4 mos, Sydney, New South Wales, Australia, • Embarked on an intellectual odyssey during my Research Internship at UNSW, Sydney, led by the erudite mentors Dr. Sebastian Sequoiah-Grayson and Carol Besseling. • Delved deep into the realm of Kantian ethics, where I meticulously crafted a comprehensive Literature Review, unearthing the intricacies of Automated Kantian Ethics and scrutinizing Rule-Based Kantian Ethics for its potential to automate moral decision-making. • Ventured into uncharted philosophical waters, exploring the application of Dyadic Deontic Logic to automate non-monotonic logic, delving into practical contradiction interpretations of the categorical imperative. • Undertook the ambitious task of formalizing the Formula of Universal Law (FUL) for categorical imperatives using Dafny, a system inspired by Harvard's pioneering work. In this endeavor, I discerned that while progress had been made, the path to achieving a truly Pure Kantian Moral Agent remained elusive. • Concluded this enlightening research journey with a two-week discourse, pondering the risks and opportunities inherent in creating a Kantian Automated Machine, a concept at the intersection of ethics and technology. • My ultimate insight from this profound exploration was the paramount importance of ensuring that an Automated Agent upholds the consistency and integrity of the ethics embedded within its programming, ultimately fostering the development of an Autonomous Moral Agent (AMA). This experience has not only expanded my horizons but also ignited a passion for ethical technology that I carry forward into my future endeavors illuminating the path to forging ethically sound, AI-driven systems, and paving the way for a future where technology seamlessly aligns with moral imperatives.

Virtual Reality and Digital Twin Developer Intern at NASA - Minex Vertically Integrated Project |, Feb 2021 - May 2021 · 4 mos, Sydney, New South Wales, Australia, • Only Technical Member from a Front line Team of 6 students to create a highly ambitious Virtual Reality and Digital Twin Simulation for Off Earth Mining of water and ice on the Moon (based on the recent scientific evidence for the presence of a high quantity of Ice, liquid, and hydrogen on the Moon) and utilizing these resources by effectively mining them and creating the Moon as a refueling station for NASA to ease off the high-intensity and far-off missions. The overall aim for the creation of the Virtual reality and Digital Twin simulation for NASA is to plan effectively the cost, resources, and materials required for the mission and to simulate the entire mission before proceeding to concoct the supplies for a real mission thereby saving the assets and cost before actually manifesting the mission and trying it out in a highly technical and effective setting beforehand. • Conducted an in-depth Literature Review of the existing solutions to the problem and found the gap in terms of the technical requirements of the project and proposed an abstract for the Paper to be presented with a team of Aerospace, mining, and electrical engineers at the "Future of Mining" Conference in Perth around December 2021. • Proposed the utility of OpenGL, C#, and unity for the creation of the VR Simulation on Oculus Rift for off-earth mining and created an algorithmic estimation using the Mini-Max Algorithm and an optimization model called WEM3 for the calculation of cost and resources in real-time based on graph embeddings and connected with a UNSW researcher at NASA Jet Propulsion laboratory to create the VR simulation firsthand. • Used the Open-cut mining and Longwell Top Coal Caving by UNSW as an inspiration for crafting the VR on the Oculus Rift. • Unfortunately, I had to step away from the role after May 2021 due to extremely complicated health circumstances, though I was originally committed through the end of 2021.

Education

Bachelors of Science, Computer Science, UNSW

High School Diploma, Physics, Chemistry, Mathematics, Computer Science, English, Music, The Asian School, Dehradun, Uttarakhand, India

Public links

github: https://github.com/ChasingTheDream369