name: Bhavesh Kumar Siddamshetty
handle: Atomix2402
role_focus: AI Engineer / Applied ML Engineer
build_mode: practical AI systems, not just notebooks
core_stack: Python, FastAPI, Pydantic, Docker, ChromaDB, LLM APIs, ML libraries
specialties:
- Retrieval-Augmented Generation
- LLM agents and tool-using workflows
- Vector search, BM25, reranking, citations, evaluation
- Workflow automation with APIs
- NLP, classification, computer vision, and explainable AI
current_status: building AI systems that are useful, secure, and explainable
open_to: AI Engineer roles, RAG/LLM internships, automation projects, applied AI collaborations
|
|
availability.signal = "open to AI/ML roles, internships, freelance builds, and strong project collaborations"
preferred.work = "RAG systems, LLM apps, automation pipelines, backend AI services"
- Secure enterprise RAG systems with access-aware retrieval, citations, and evaluation.
- LLM agents that connect APIs, databases, documents, and workflow tools.
- Automation pipelines that turn emails, tickets, and messy inputs into structured actions.
- Better project documentation, demos, and interview-ready technical explanations.
|
I build RAG pipelines that retrieve the right context, apply access control, cite sources, and reduce hallucination. Keywords: RAG, embeddings, ChromaDB, BM25, RRF, reranking, RBAC, citations |
I design LLM-powered workflows that parse, classify, route, and act across APIs and real-world tools. Keywords: agents, triage, Gmail API, Notion API, GitHub Actions, structured extraction |
|
I have explored NLP, sentiment analysis, classification, computer vision, YOLO, TensorFlow, and explainability. Keywords: NLP, CV, XAI, notebooks, evaluation, model experimentation |
I like making AI work accessible through APIs, CLIs, Docker setups, docs, and repeatable evaluation scripts. Keywords: FastAPI, Typer, Pydantic, Docker, pytest, CLI, API design |
|
Secure enterprise RAG over CSV, JSON, SQLite, Markdown, policies, and reports. What stands out:
|
Multi-domain support triage agent for classifying tickets and generating responses. What stands out:
|
|
Automated tracker that scans Gmail, parses job emails with AI, and syncs updates into Notion. What stands out:
|
Natural-language interface for querying SQL-style data through an LLM-driven agent layer. What stands out:
|
|
Search-oriented LLM project exploring retrieval and language-model powered search workflows.
|
Computer vision experiment using YOLOv8 for object detection workflows.
|
[01] Build secure RAG systems with access-aware retrieval
[02] Design LLM agents that can reason over APIs and tools
[03] Improve reliability with citations, traces, evals, and guardrails
[04] Convert AI experiments into usable workflows and interfaces
I build AI systems where retrieval, reasoning, automation, and reliability meet practical software engineering.