Sainesh Nakra
Experienced Software Engineer
About Me
Hey, I'm Sainesh.
A Software Engineer at Google, having recently completed a Master of Science in Computer Science at Northeastern University, Boston — specializing in machine learning and distributed systems.
At Northeastern I was a Graduate Teaching Assistant for Discrete Mathematics, and a Research Assistant in two labs — working with Prof. Konrad on smart contract auditing using generative AI, and with Prof. Mallesh on machine learning techniques for VR video compression. I also advise MIT students building AI products as a Course Advisor under Prof. Ramesh Raskar.
Previously, I worked in software development at Bosch, Honda, and NEC, with my code impacting over a million users. I've also won two hackathons and reached the finals in three others.
running my code
running my code
...and, surprisingly, they still work.
I'm passionate about solving complex problems with technology. If you're interested in working together, please reach out!
My Skills
Work & Education
Work
work
Software Engineer
↪ San Jose, CA
- Developing mobile data download infrastructure to deliver feature and on-device LLM updates without APK releases, enabling checkpoint-based rollout of fine-tuned local models across 13 countries and reducing rollout time from 3 months to 24 hours.
work
MathWorks
Software Engineer
↪ Boston, MA
- Collaborated on MATLAB-integrated Copilot by shipping TypeScript editor features for LLM-backed code completion, chat, and refactoring in developer workflows.
- Developed RAG-style context grounding for an AI-enabled code editor using file and symbol context, adding guardrails and fallbacks to improve response relevance and reliability.
work
Patient First.AI
Software Engineer
↪ Boston, MA
- Built and launched a vaccination-compliance platform for 10+ universities using Fastify/Node.js APIs and React dashboards, backed by PostgreSQL to manage 50,000+ records and enforce 100+ policy rules.
- Automated real-time verification of 100,000+ documents by designing an OCR and logic-based parser with >95% field-extraction accuracy and an AI-powered pipeline achieving 91% recognition with sub-30-second processing.
work
MIT, under Prof. Ramesh Raskar
Course Advisor
↪ Cambridge, MA
- Advising students in the 'Foundations of AI Ventures' course on developing AI-based products.
- Providing mentorship on codebases and product development for student-led startup projects, including personalized sessions to resolve technical challenges.
work
Northeastern University, under Prof. Mallesh Dasari
Research Assistant
↪ Boston, MA
- Conducted research on 3D video compression using Neural Video Codecs, focusing on spatial video efficiency.
- Worked with diffusion models to enhance video processing and compression techniques.
work
Khoury College of Computer Science, Northeastern University
Research Assistant
↪ Boston, MA
- Led the development of ML-powered tools for streamlining smart contract audits, enhancing efficiency by 80%.
- Integrated established tools with language model-driven insights, resulting in a 30% increase in audit accuracy and reducing processing time from days to minutes.
work
Bosch Global Software Technologies
Software Engineer
↪ Bengaluru, India
- Led a 3-member team to develop Radar List Handling for Adaptive Cruise Control and Emergency Braking by integrating radar and camera signals with C++ and Python, applying computer vision and ML-based validation, earning the 2022 Agile Award and increasing diagnostic accuracy by 15%.
- Implemented 7+ web features for Audi and Volkswagen cloud platforms in TypeScript on AWS, increasing system reliability by 25% and enhancing safety features for over 1 million users.
work
Bosch Global Software Technologies
Software Engineer Intern
↪ Bengaluru, India
- Integrated Bosch's Central Secure Gateway software into PACCAR CSG2 SW for major truck brands using C++ on Linux, improving firewall efficiency and supporting secure vehicle communication across 185,900+ vehicles.
work
Honda Cars India Pvt Ltd
Software Engineering Intern
↪ India
- Redesigned the UI of Honda Cars' mobile app, resulting in a 15% increase in user engagement and satisfaction.
work
NEC Technologies Pvt Ltd
Software Engineering Intern
↪ India
- Contributed to back-end development using RabbitMQ for real-time messaging in a government defense project, improving load-handling capacity by 400%.
work
VIT Innovation Labs
Research Assistant
↪ Vellore, India
- Led rapid prototyping sessions for VR depth perception studies, conducting over 200 trials to gather patient data and refine product features.
Education
edu
Northeastern University
Master of Science in Computer Science
↪ Boston, MA
- Relevant Coursework: Software Engineering, Scalable Distributed Systems, Algorithms, Programming Design Paradigms, Natural Language Processing, Machine Learning.
edu
Vellore Institute of Technology
B.Tech in Electronics and Communication Engineering
↪ Vellore, India
- Relevant Coursework: Operating Systems, Computer Networks, Web Development, Database Management Systems.
Featured Projects
ScriptCut — AI Video Editor
Founded ScriptCut, an AI video editor powered by a fine-tuned LLM, deployed via a Dockerized Next.js + FastAPI stack. Edit videos by editing the transcript: precise control over content, automatic blooper removal, and text-edits that reflect in the video itself. Reduces video editing time by 80%.
Tutu — Educational Song Generator
Built Tutu, an AI-based platform that creates educational songs based on user-selected topics. Tutu uses NLP and audio synthesis to generate topic-focused songs, helping make learning more enjoyable. Users can easily enter their topics and receive customized songs to support learning and retention.
Automated Data Analyst
Automated Data Analyst is an AI-driven platform that simplifies data analysis for users. By uploading a CSV sheet and providing a brief description of the data, users can automatically perform trend analysis, derive business insights, and receive actionable conclusions. The platform leverages advanced data processing and machine learning techniques to assist with everything from exploratory data analysis to high-level business inferences, making complex data accessible to everyone.
cAIro — AI-Powered Digital Marketing Platform
Developed cAIro, an AI-driven platform designed to streamline digital marketing efforts for businesses. cAIro provides tools for creating professional landing pages, generating high-quality marketing content, and automating lead generation, allowing businesses to focus on growth. With its time-saving automation features and access to expert-level marketing resources, cAIro helps businesses enhance their online presence and reach a broader audience.
AI-Powered Smart Contract Auditor
Led the development of DefiGuard, an innovative smart contract auditing tool that uses advanced language models to streamline the auditing process for Ethereum-based contracts. This solution significantly improved efficiency, cutting down manual auditing work by 80% and ensuring thorough and accurate audits.
Insurance GPT
Developed an AI-powered assistant for navigating health insurance services. The application allows users to upload their insurance policy PDFs to discover preventive and free services covered by their insurance. It also helps users find booking locations based on their policy and location inputs. The website features a stunning and interactive user interface with advanced CSS styling for an enhanced user experience.
Homewrecker.ai — Textual Analysis for Disputes
Homewrecker.ai uses advanced textual analysis to help couples resolve their heated arguments. By uploading conversations, the platform employs AI to analyze the dialogue, identify key emotions, and determine who is 'right' in the dispute. The result? No more endless arguments — just data-driven clarity.
GetMeAJob — Job Application Tracker
Developed a web application to help users track job applications and stay organized during the job hunt. GitHub is private — please visit the website instead.
Netflix Clone (SNSTREAM)
Used React to recreate the Netflix UI with real-time data updates for shows, using API endpoints to ensure fresh content each time.
Tavern Social
Developed an informative website for a blockchain-based social media platform currently in the development phase.
E-Commerce Website UI
Designed the frontend experience for an e-commerce store landing page using HTML, CSS, JavaScript, and jQuery.
LinkedIn Clone Using React
Developed an interactive LinkedIn clone with advanced authentication, allowing seamless user functionality for posting content and managing profiles.
Game Projects
Road Kill
Road Kill is an engaging and dynamic racing game that integrates blockchain technology to offer a thrilling gaming experience while ensuring secure and transparent leaderboard management. Players navigate through various levels, dodge obstacles, collect rewards, and customize their cars with different skins. The game's unique blend of fast-paced action and blockchain integration offers a novel approach to racing games.
Arcade Snake Game
Recreated the popular snake video game with a cleaner UI using HTML, CSS, and vanilla JavaScript.
The 2048 Game
Redesigned the UI for the popular 2048 game to make it ad-free and visually engaging. Implemented using JavaScript and made open-source for customization.