Skip to content
View Ashindustry007's full-sized avatar
📕
Learning
📕
Learning
8 followers · 10 following

Achievements

Achievement: Pull Sharkx2Achievement: YOLO

Achievements

Achievement: Pull Sharkx2Achievement: YOLO

Highlights

  • Pro

Block or report Ashindustry007

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
Ashindustry007/README.md

MSCS student at UC San Diego and AI Engineer building intelligent systems through computer vision and agentic architectures. Focus: AI agents engineering, AI agents plus computer vision, and systems for AI agents.

About me. Shaping The Future Of Ai-Powered Creation. I am an AI Engineer pursuing my MSCS at UC San Diego, bridging theoretical research and scalable, real-world systems. My focus sits where computer vision meets agentic AI. With a foundation from OUTR (B.Tech), I have built intelligent pipelines for hard problems, from cancer-risk assessment via spatial transcriptomics to agentic automation of pre-silicon chip design. Right now I am building agents that act on real systems and can show their work: a cited legal-case digest, vision-grounded reasoning over scene graphs, and human-approved browser automation. Coder, artist, traveler. MSCS at UC San Diego.

Current Focus

Ai Agents Engineering. Agents that plan, call tools and cite their sources, kept in check by verified outputs and human-in-the-loop guardrails. Seen in: Case Pulse · Qualcomm.Ai Agents + Cv. Giving agents eyes: vision-language models reasoning over verified scene graphs, grounded in hands-on segmentation research with self-supervised vision transformers. Seen in: Lrtsg · Boolean Lab.Systems For Ai Agents. The plumbing around the model: WebSocket bridges, secure backends, streaming and retrieval that let agents run on real, sensitive systems. Seen in: Qualcomm · Case Pulse.

Past Experience

Past experience. Ai Engineer Intern, Qualcomm: AI-assisted browser automation for pre-silicon chip-design tools: engineers drive internal web portals in natural language. Built and compared an LLM bridge and a stateful agentic Chrome extension, with per-page guardrails and mandatory user approval. Highlight: 53 browser-automation playbooks authored. Ai Research Intern, Boolean Lab · Ucsd Health: Self-supervised computer vision for colon-tissue analysis: DINOv3-7B Vision Transformers fine-tuned with LoRA (75% less memory) plus a Mask2Former head, cutting manual annotation time by 90%. Highlight: 81% miou in crypt-level segmentation. Software Engineer, Open Financial Technologies: Two years on a neobanking platform after an internship: multi-modal invoice AI, payroll and payment-gateway work, and RBI-compliant risk controls protecting 50K+ users. Highlight: 92% of invoices auto-converted. Deep Learning Research Intern, Pucho Digital Healthcare: Deep learning for healthcare: in-silico mRNA vaccine design and explainable computer vision for disease-cell detection in MRI scans, using XAI to justify predictions. Highlight: 96.55% sensitivity in disease-cell detection. Data Science Intern, Analytics Labs: Predictive and anomaly-detection models across epidemiology, finance and predictive maintenance, including 89%-accurate COVID-19 spread forecasts from a hybrid statistical and deep-learning model. Highlight: 98.5% recall on credit-card fraud.

Selected Achievements

Winner: GDG Challenge at DataHacks 2026, with Allerion, an AI-powered, species-specific pollen forecasting system. Competed against 100+ teams (450+ participants), judged by 80+ judges and mentors.

Achieved world rank 96th in TCS CodeVita Season 10.Ranked among the top 5 in the AWS Generative AI Hackathon, conducted by AWS at Open Financial Technologies.Kaggle Expert in the Notebooks category. View profile.Published paper: Odia Handwritten Numeral Recognition, a hybrid modelling approach, at ICECA 2021 (IEEE). Read the paper.

Repo Catalogue

Case Pulse, Swans 2026 · Law-Di-Gras Hackathon. Cited case digestion for personal-injury firms. Reads Clio read-only and ties every fact to its exact source span, with 147/147 citations verified at $1.10 per full case.Allerion, Datahacks 2026 · Gdg Challenge Winner. Species-specific pollen forecasting that fuses iNaturalist citizen-science data, the Google Pollen API and a phenology engine into 14-day, neighborhood-level forecasts, with Gemini 2.5 Flash agents for advisories and photo-based plant ID.

Lrtsg, Language Reasoning With Scene Graphs. Zero-shot agentic pipeline for visual reasoning: six agents plan, narrate, build, localize, audit and answer over a verified Scene Graph, using ReAct/CodeAct tool use with Qwen3-VL or Gemini.Safe, Smart Analytics For Fire Emergencies. Wildfire intelligence platform from the Reboot the Earth hackathon: Rothermel fire-spread simulation on 3D terrain, plus an LLM assistant that fails over across Groq, Gemini, OpenAI/Mistral and a local fallback.

Musicnet, Generative Ai For Music. Beethoven-style music generation with an LSTM and embedding model trained on 157 MIDI pieces across 130 note classes, rendered back to audio with music21.Retinal Vessel Segmentation, Diabetic Retinopathy · Drive. Pixel-level segmentation of retinal images on the DRIVE dataset, classifying vessel versus non-vessel pixels with a U-Net as groundwork for early diabetic retinopathy detection.

Final call: let’s build intelligent systems together.

Portfolio Linkedin Email Kaggle

Pinned Loading

  1. LRTSG---Language-Reasoning-Task-with-Scene-Graph LRTSG---Language-Reasoning-Task-with-Scene-Graph Public

    Given an image I and a natural language question q, output an answer a to an arbitrary relationship reasoning question with no task-specific training, fine-tuning, or labeled examples. Scope & Assu…

    Python

  2. DataHacks26-Allerion DataHacks26-Allerion Public

    Allerion is the world's first AI-powered, species-specific pollen forecasting system. It fuses crowdsourced citizen-science observations with atmospheric APIs to give allergy sufferers personalized…

    Python

  3. SAFE SAFE Public

    SAFE is a high-fidelity wildfire intelligence platform designed for real-time fire behavior prediction, regional risk assessment, and emergency response coordination. It utilizes the Rothermel Surf…

    HTML 1

  4. Vocal-Biomarker-ICBHI-final-database Vocal-Biomarker-ICBHI-final-database Public

    The automatic analysis of respiratory sounds has been a field of great research interest during the last decades. Automated classification of respiratory sounds has the potential to detect abnormal…

    Jupyter Notebook 3

  5. Case-Pulse Case-Pulse Public

    LAW-DI-GRAS Applied AI Hackathon

    Python

  6. codeFafnir/sdx_hackathon_404_not_found codeFafnir/sdx_hackathon_404_not_found Public

    Hackathon repo

    Python