AI ENGINEER · OPEN TO WORK · LUCKNOW, INDIA
Building practical AI systems with LLMs, RAG, agents, and automation.
I’m Asheesh Ranjan Srivastava, an AI Engineer focused on turning complex problems into useful, testable products. I combine AI engineering, product thinking, and systems design to build tools that help people learn, work, and make better decisions.
4 years
Total professional experience
400+
Professionals trained through nonprofit education programs
5 states
Nonprofit education programs supported across India
AI Engineering
Current portfolio and career focus
ENGINEERING FOCUS
From prototypes to reliable AI product workflows
LLM applications
Designing useful product experiences around language models, structured outputs, and multi-model orchestration.
RAG systems
Building retrieval pipelines that connect models to curated knowledge, citations, and evaluation criteria.
Agents & automation
Orchestrating tools, decisions, and workflows with agents, n8n, APIs, and human review.
SELECTED WORK
AI systems built around real learning and workflow problems
The featured case study reflects currently verified public repository details. Demos, visuals, outcome metrics, and detailed ownership will be added as evidence is confirmed.
FEATURED PROJECT · PAUSED
Quest & Crossfire Curriculum Studio
An AI-powered curriculum-generation system for Python education, combining a five-agent sequential pipeline, adaptive teaching modes, retrieval-augmented generation, and automated quality gates.
- Five-agent workflow: research, synthesis, structure, compilation, and narrative enrichment
- Coach, Hybrid, and Socratic teaching modes
- FastAPI, LangGraph, LangChain, PostgreSQL, pgvector, and Streamlit
- Quality gates, tracing, authentication, and self-improving RAG workflows
My current contribution: architecture design and debugging guidance. Detailed ownership and AI-assistance attribution will be added after review.
Live demo URL: to be added when the project resumes.
PROJECT VISUAL PLACEHOLDER
Architecture diagram or product screenshot to be added
EVIDENCE PLACEHOLDER
Outcome metrics and evaluation results: to be decided.
Additional projects
PROJECT 2 · TO BE SELECTED
Next flagship case study
Reserved for a public, demonstrable project that broadens the portfolio beyond the paused Aethelgard ecosystem.
Selection criteria: working demo, public code, clear ownership, and measurable engineering evidence.
MULTI-MODEL · COST OPTIMIZATION
Text Summarization Workflow
A multi-model summarization workflow designed to preserve useful output quality while reducing model cost.
Result placeholder: Cost and quality evaluation to be measured and documented.
Let’s build useful AI systems.
I’m open to AI Engineer opportunities where practical engineering, product judgment, and responsible AI collaboration matter.