Sugar-AI (Memory Optimization)
Contributed to Sugar Labs by eliminating double-initialization of RAGAgent through lazy loading and FastAPI startup event dependency injection, reducing startup RAM consumption by approximately 50%.
Backend Engineer
Backend engineer and AI/ML student building Django and FastAPI services, RAG-based applications, local AI systems, and computer vision solutions.
Blockchain Research Lab
Jan. 2024 – Present · Ghaziabad, India
Architected blockchain data-indexing services using Django and REST APIs; built a WebSocket streaming pipeline for live event monitoring, eliminating polling overhead. Optimized database queries through profiling and index tuning, reducing average API response latency by approximately 30% under load testing.
TEDx AKGEC
Managed technical and logistics teams.
geeksecurity
Founded an educational AI and cybersecurity community with more than 22,000 active followers.
NYAYAGPT-AI
1 yr 1 mo · Gurugram
Building NyayaGPT — an AI-powered legal intelligence platform that converts real-world crime descriptions into structured FIRs application and maps them to relevant BNS/BNSS sections.
Blockchain Research Lab
1 yr 9 mos · Ghaziabad, Uttar Pradesh, India
Blockchain Research Lab
3 mos · Ajay kumar garg Engineering college
Fiverr
2 yrs 8 mos
I build "set it and forget it" content pipelines using Python and AI. My automated systems connect to live data, write unique articles with Gemini 2.5 Pro, create stunning images with Imagen, and publish everything directly to your WordPress site with full SEO. Perfect for auto-blogs, news sites, or any platform needing high-volume content. Skills: Automation: Python, FastAPI, WebSockets AI: Gemini, Imagen, LLM Integration, RAG Based Web: WordPress REST API
AKGEC IDEA Lab
1 yr 9 mos · Ghaziabad, Uttar Pradesh, India
SECUREVERSE
4 mos
• Developed and maintained backend applications using Python and frameworks such as Django/Flask • Designed and implemented RESTful APIs to support web and mobile applications • Built automation scripts to streamline internal workflows, reducing manual effort and increasing operational efficiency
Markup Media
2 mos · Noida, Uttar Pradesh, India
Bachelor of Technology, Artificial Intelligence and Machine Learning
2023 – 2027
Bachelor of Technology - BTech, Artificial Intelligence and Machine learning
Sep 2023 - Aug 2027
Bachelor of Technology
2023 - 2027
Intermediate, Physics and Mathematics
Apr 2022 - May 2023
Contributed to Sugar Labs by eliminating double-initialization of RAGAgent through lazy loading and FastAPI startup event dependency injection, reducing startup RAM consumption by approximately 50%.
Designed a Creative Persona system prompt architecture for offline 135M-parameter LLMs after evaluating and rejecting a two-pass reviewer loop, reducing hallucination risk through a safe storytelling companion scope.
Developed a RAG system mapping natural-language crime descriptions to relevant BNS/BNSS Acts. Reduced hallucinations by approximately 50%, achieved approximately 85% retrieval accuracy over 1,500+ indexed law sections, reduced inference latency from approximately 3.8 seconds to 2.1 seconds, and automated FIR draft generation in simulated workflows.
Built a multi-step Planner–Executor–Reflector LLM agent that converts text prompts into formatted Word documents using schema-constrained JSON planning. Added bounded self-checking, defensive JSON parsing, and explicit assumption recording for ambiguous requests.
Developed a computer vision gesture recognition system using OpenCV and MediaPipe for hands-free media navigation, achieving approximately 25–30 FPS, approximately 90% recognition accuracy under normal lighting, and under 150 ms interaction latency.
geeksecurity
TEDxAKGEC
Jan 2025 - Apr 2025 · 4 mos