Skip to content
View suhailult777's full-sized avatar
5 followers · 5 following

Achievements

Achievement: Pair ExtraordinaireAchievement: Pull Sharkx2Achievement: YOLOAchievement: Quickdraw

Achievements

Achievement: Pair ExtraordinaireAchievement: Pull Sharkx2Achievement: YOLOAchievement: Quickdraw

Highlights

  • Pro

Block or report suhailult777

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. 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
suhailult777/README.md

Mohd Suhail Idrisi

AI engineer · RAG pipelines, LLM agents and full-stack Next.js apps · Hyderabad, India

I'm a 2026 B.Tech Computer Science graduate. I've shipped generative-AI and full-stack features in production: the RAG data pipeline behind a GST-compliance platform, and customer-facing features for an AI content-monetization SaaS. I like taking a rough idea and turning it into something people can use, and I'm open to full-time roles and freelance AI chatbot/RAG work.

Featured projects

Project What it is Stack
AI-multi-code-agent Autonomous coding agent: Gemini plans and writes whole projects through a tool-calling loop (think → act → observe), and a second Codestral agent reviews the output. Streams progress to a web UI over SSE. Node.js, Express, Gemini, Mistral, PostgreSQL
MedBrain-0.5B 0.5B-parameter medical language model I fine-tuned (LoRA, JAX/Flax) and published on Hugging Face. Includes an inference script and a minimal ChromaDB RAG example. JAX/Flax, PyTorch, Transformers
PocketClaw My B.Tech major project: Android AI chat app that runs in the cloud (NVIDIA NIM) or fully offline with on-device GGUF models via llama.rn, with web-search tool calls. React Native, Expo, TypeScript, llama.rn
COREP Reporting Assistant RAG prototype that fills a banking regulatory template (PRA COREP C 01.00) with values, reasoning and citations; Pydantic-schema output plus arithmetic validation and Excel export. Python, Gemini, embeddings, Pydantic, Streamlit
Time-Travel Evals Hackathon project: evaluates AI agents across perturbed "alternate timelines" (crashes, latency spikes) and scores robustness and recovery, with a Streamlit trace viewer. Python, OpenAI-compatible APIs, Streamlit
Two-stage Recommender Two-tower retrieval + FAISS + wide-and-deep ranker on MovieLens 100K behind FastAPI, with measured Recall/NDCG against a popularity baseline. TensorFlow, PyTorch, FAISS, FastAPI

MedBrain-0.5B on Hugging Face

huggingface.co/suhailult777/MedBrain-0.5B: a 0.5B medical LLM that gives structured answers for triage, clinical handoffs and patient education.

  • LoRA (rank 16) fine-tune on medical-instruction-100k, trained in JAX/Flax and exported to PyTorch.
  • 5,800+ downloads.
  • Research artifact only; not for clinical use.

Experience

AI Developer Intern, ApexAgiSolutions (Apr 2026 – present)

  • Building the data and RAG pipeline behind TaxApex, a GST-compliance platform.
  • Invoices and ledger records go into PostgreSQL, then get cleaned, chunked, embedded and indexed so the assistant answers grounded in source records.
  • Built the REST APIs for the chartered-accountant dashboard: GSTR-2B reconciliation, ITC gap analysis, and GSTIN and HSN validation.

Full-Stack Developer, PayPerCrawl (Sep 2025 – Apr 2026)

  • Shipped production features for an AI content-monetization SaaS in Next.js, React, TypeScript and Tailwind.
  • Built REST APIs, auth flows and database integrations, and debugged live issues from logs back to the failing request.

Stack

  • AI / ML: Python, LangChain, Hugging Face Transformers, PyTorch, JAX/Flax, OpenAI and Gemini APIs, RAG, embeddings, ChromaDB, Pinecone
  • Backend: Node.js, Express, REST APIs, PostgreSQL (Neon), MongoDB, Drizzle, SQL
  • Frontend: React, Next.js, TypeScript, Tailwind CSS, React Native (Expo)
  • Tools: Git, Vercel, Streamlit, FastAPI

Contact

Pinned Loading

  1. MedBrain-0.5B MedBrain-0.5B Public

    0.5B medical LLM (LoRA fine-tune, trained in JAX/Flax, exported to PyTorch). 5.8k+ downloads on Hugging Face.

    Python

  2. AI-multi-code-agent AI-multi-code-agent Public

    Autonomous coding agent: Gemini plans and writes projects through tool calls, Codestral reviews them. Node.js, Express, SSE UI.

    JavaScript

  3. pocketclaw pocketclaw Public

    B.Tech major project: React Native/Expo AI chat with NVIDIA NIM cloud and fully offline on-device GGUF inference (llama.rn).

    TypeScript

  4. codex-community-hackathon-hyd-time-travel-eval codex-community-hackathon-hyd-time-travel-eval Public

    Time-Travel Evals: stress-test AI agents across perturbed 'alternate timelines', with robustness metrics and a Streamlit trace viewer.

    Python

  5. corep-reporting-assistant corep-reporting-assistant Public

    LLM + RAG prototype that fills PRA COREP C 01.00 with cited, schema-validated values: Gemini, Pydantic, Streamlit, Excel export.

    Python

  6. nexvia-aiml03-recommender nexvia-aiml03-recommender Public

    Two-stage recommender on MovieLens 100K: two-tower retrieval + FAISS + wide-and-deep ranker behind FastAPI, measured Recall/NDCG.

    Python 1