ATHARVA
JOSHI
SOFTWARE · ENGINEER
The Engineer Behind the System
I'm a software engineer who builds systems that work — not just code that compiles. My approach bridges the gap between deep technical engineering and production-grade software delivery.
From training computer vision models on satellite imagery to building autonomous drone landing systems, from designing full-stack platforms to implementing real-time collaborative systems — I operate across the entire engineering spectrum.
My philosophy is simple: every technical claim should be backed by something real. This portfolio itself is a testament to that principle — a full-stack application demonstrating the very skills it showcases.
Featured Projects
Autonomous Drone Mission Control
ISRO IRoC-U 2025 autonomous UAV system for GPS/GNSS-denied navigation and autonomous mission execution — AIR 1 among 500+ teams.
Adversia — AI Payment Security Lab
AI Red Team ↔ Blue Team payment fraud detection and adversarial security system built for the Mastercard Hackathon.
SovereignGuard — Secure LLM Data Pipeline
Secure and sovereign LLM data pipeline that lets organizations use LLMs with proprietary data without exposing it to external model providers — B.E. capstone.
Green Model Advisor
Carbon-aware AI inference router that dynamically selects LLM/inference routes based on latency, accuracy, computational cost, and estimated CO₂ emissions.
Satellite Air Pollution Analytics
ISRO Bharatiya Antariksh AI/ML Hackathon 2025 end-to-end ML/data pipeline for analysing satellite and environmental data for air-pollution analysis — Top 28 / 8,744 teams.
CollabIQ
AI-powered SaaS project management platform with AI-assisted task management, documentation, collaborative coding, and analytics dashboards.
SHETISAKHA
Agri-Bot control system — an Android 14-compatible farming robot control app with live soil monitoring and autonomous path replay.
Technical Proficiency
</>Software & Backend
AIAI/ML & GenAI
DBData & Big Data
☁Cloud & Infrastructure
RSRobotics & Systems
CSCore Computer Science
Professional Journey
AI / Software Engineering
Working on secure and sovereign LLM data pipelines for enterprise AI systems, focusing on protecting proprietary company data throughout the AI lifecycle.
- Developing data pipelines for ingestion, preprocessing, classification, and secure preparation of enterprise data for LLM applications.
- Working on mechanisms to detect and protect sensitive and confidential information before it reaches LLM systems.
- Exploring data isolation, governance, and sovereignty to prevent unauthorized cross-organization data exposure.
- Contributing to secure LLM/RAG pipelines designed to reduce the risk of sensitive information leakage.
Finance & Project Management Lead
Leading finance and project management alongside technical planning across a 30-member robotics team competing in national-level robotics competitions.
- Managing project planning and milestones.
- Coordinating task pipelines across technical teams.
- Tracking integration milestones.
- Supporting architecture and technical coordination.
- Managing budgeting and procurement.
- Planning resource allocation.
- Coordinating between different technical teams.
- Supporting testing, debugging, documentation, and integration.
- ★AIR 1 — ISRO IRoC-U 2025
- ★AIR 9 — DD Robocon India 2026
- ★Best Technical Report Award
- ★100/100 in Stage 1 and Stage 2 at DD Robocon India 2026
Open Source Contributor
Contributing to open-source projects through repository issues, codebase exploration, and validated fixes using issue-based development workflows.
- gVisor — worked through repository issues, understood the existing codebase, implemented fixes, and validated the changes.
- eksctl (official CLI for Amazon EKS) — worked on issues with corresponding code and test changes.
- SoL-Pi — contributed to the project and its development workflow.
- Built comfort with reading unfamiliar codebases, Git/GitHub workflows, issue tracking, debugging, and writing validated fixes.
Computer Vision / ML Intern
Worked on computer-vision and machine-learning workflows involving large-scale image datasets.
- Built batched and parallelized Python ETL pipelines for image preprocessing and data preparation.
- Added REST-controlled orchestration to make the data-processing workflow easier to integrate with other services.
- Reduced image-preprocessing latency by 35%, measured through pipeline throughput benchmarks.
- Worked on YOLO model fine-tuning, custom data augmentation, and MLflow experiment tracking.
- Improved YOLO detection mAP by 18% over the baseline on held-out validation data.
- Focused on reproducibility by developing unit-tested ML workflows and reproducible builds.
Research Lead Intern — AI/ML
Led a 6-member team working on Brain-Computer Interface (BCI) research as an AI/ML Research Lead.
- Led the team through the complete design → build → evaluate ML workflow.
- Worked with EEG signals and BCI classification.
- Developed an ML classification system that achieved 82% accuracy on held-out test data.
- Established a reusable ML research methodology that was adopted across the team.
- Worked on research, experimentation, evaluation, documentation, and team coordination.
Web Developer Intern
Worked as a Web Developer Intern and contributed to the development of the company's official website.
- Designed and developed the official website for the corporate fitness brand.
- Implemented responsive UI and user-friendly website features.
- Worked on aligning the website's design and functionality with the company's branding and promotional requirements.
- Gained practical experience in translating business requirements into usable web interfaces.
Web Developer Intern
Worked on full-stack web development during the diploma final year.
- Developed dynamic and user-friendly web interfaces.
- Worked with HTML, CSS, JavaScript, Bootstrap, and PHP.
- Gained practical experience in server-side development and database integration.
- Worked on creating web applications with functional frontend, backend, and database components.
- Developed an understanding of how web applications move from UI development to server and database implementation.
Academic Background
B.E. Computer Engineering
Lateral entry (2nd year, via Diploma) pursuing Computer Engineering with a focus on software systems, AI/ML, and cloud engineering.
- ▸Lateral entry directly into 2nd year via Diploma in IT
- ▸Actively involved in PICT Robotics — AIR 1 in ISRO IRoC-U 2025
- ▸Core CS: Data Structures, OOP, DBMS, Operating Systems, Networks, System Design
Diploma in Information Technology
Diploma in Information Technology building a strong foundation in programming, databases, and web technologies.
- ▸Strong foundation in DSA, OOP, and DBMS fundamentals
- ▸Built hands-on programming and web development skills
Recognition & Impact
ISRO IRoC-U Robotics Challenge
Ranked 1st nationwide (AIR 1) among 500+ teams for an autonomous UAV with GPS-free navigation, LiDAR-based terrain mapping, and computer-vision landing-zone detection.
DD Robocon
Finished AIR 9 nationally with 100/100 in Stage 1 & Stage 2 and the Best Technical Report Award.
ISRO Bharat Antariksh Hackathon
Finished Top 28 out of 8,744 teams with an ML-based air quality index (AQI) prediction pipeline.
Adversia — AI Payment Security Lab
Built an AI Red Team ↔ Blue Team payment fraud-detection system with 92/92 unit tests and 14/14 end-to-end tests; adversarial recall improved by 13.5 pp and F1 from 0.357 to 0.519.
Finance & Project Management Lead
Leading a 30-member robotics team — owns project planning, budgeting, and integration milestones; key outcomes include AIR 1 (IRoC-U 2025) and AIR 9 (DD Robocon 2026).
Get in Touch
OPEN TO SOFTWARE ENGINEERING OPPORTUNITIES — I'm always open to discussing new projects, creative ideas, or opportunities to be part of your vision.