Skip to content
View hoomanesteki's full-sized avatar

Highlights

  • Pro

Block or report hoomanesteki

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.

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
hoomanesteki/README.md

Website  LinkedIn  YouTube  Instagram  Location

I take AI systems from prototype to production: grounded RAG, multi-agent systems, and the evaluation and guardrails that make them safe to ship. Safety lives in code, not prompts. Every claim carries a number.

🎓 Master of Data Science, UBC  ·  ☁️ AWS & Databricks certified  ·  📍 Open across Canada and the US

     

I work where “the model works” becomes “the system is ready to ship.”


🚀 Selected Work

Nine systems, end to end. Every card reads the same way: what it doesthe result → how it works → the stack.  Browse all nine →

Stops a code change from quietly breaking your AI.

10 / 10 planted bugs caught · $0 to run

Records the agent once, replays it offline with no API keys, and fails the build before a bad change merges.

Python OpenTelemetry SQLite GitHub Actions

case study code

Finds the cheapest mix of AI models that still gets the job right.

3.6× cheaper at the same quality

Splits an agent into planner, doer and checker, then scores all 15 combinations on the same 8,100 graded runs.

Python FastAPI Anthropic OpenAI Groq

case study code

A company assistant that answers with sources, or admits it doesn't know.

85.9% routing · 100% of hard cases escalated

One orchestrator picks a specialist across documents, governed metrics and a knowledge graph, then checks every citation.

LangGraph Qdrant Neo4j dbt FastAPI

case study code

Turns raw social conversation into customer types you can actually prove.

4 personas + 11 sub-types, each statistically validated

An LLM tags behavioural signals at scale, three clustering algorithms agree on the groups, and significance tests confirm each one is real.

Python DSPy HDBSCAN scikit-learn MLflow

case study

Reads sales calls and decides which misrouted lead earns a refund.

✅ Precision 43% → 89% on the same model

The LLM only extracts facts; plain code makes the decision, so a bad guess shows up as a visible error instead of a silent one.

Python LLM APIs XGBoost Embeddings

case study code

Ask a data question in plain English, get the chart and the answer.

91% routed correctly · 7 specialist agents

An orchestrator hands the question to the right specialist, which runs its code in a sandbox behind 40+ security checks.

FastAPI MCP Plotly pandas Docker

case study code

Notices when a model goes stale and retrains it on its own.

5-stage closed loop · zero manual scheduling

A drift metric watches live traffic and triggers retraining automatically, but a human still approves the new model.

FastAPI MLflow XGBoost Docker AWS

case study code

Search that understands what you meant, not just what you typed.

+22% better results, A/B tested

Keyword and meaning-based search run together and their rankings are fused, then a guarded assistant answers from the evidence only.

FAISS BM25 SBERT LangChain Gradio

case study code

An AI analyst that asks permission before it touches your database.

1M+ rows queried · zero writes allowed

Three approval gates before anything runs, and read-only SQL enforced at the syntax-tree level rather than by a prompt.

FastAPI Gradio MCP SQLite pytest

case study code


🧭 How I Build

🛡️ Guarded
Safety rules and approval gates live in code, not prompts.
📏 Measured
Models chosen on quality, latency, cost and failure analysis.
💸 Cost-Aware
Rules and small models handle the easy work first.
🤝 Human-Approved
Retraining is automatic. Promotion to production is not.

🛠 Toolbelt

🧵 RAG & Retrieval

LangGraph LangChain Qdrant FAISS BM25 SBERT Cohere Neo4j RRF Reranking

🤖 Agents & Safety

MCP DSPy Multi-agent routing Guardrails Human-in-the-loop Prompt-injection defence

🚀 MLOps & LLMOps

MLflow Langfuse RAGAS OpenTelemetry FastAPI Docker CI eval gates Drift monitoring

🧠 ML & NLP

PyTorch scikit-learn XGBoost Transformers Embeddings Clustering

🗄️ Data & Cloud

Spark Databricks dbt DuckDB PostgreSQL SQLite AWS Azure

📊 Statistics

Python R A/B testing Causal inference Bootstrap CI McNemar Fisher / χ² BH-FDR


🌄 Life Beyond the Laptop

Mountains, movement, and a bit of competition.


⛷️ Skiing

🎾 Tennis

🏃 Running

🐎 Polo

🥾 Hiking

🏋️ Strength

📊 GitHub Activity


def ship(problem):
    return build(problem,
        grounded=True,       # cite the evidence, or abstain
        guarded=True,        # safety in code, not in the prompt
        measured=True,       # a number behind every claim
        cost_aware=True,     # the cheapest model that still passes
        human_approved=True, # a person promotes to production
    )

Portfolio   LinkedIn

Machine Learning Engineer · Data Scientist · Master of Data Science, UBC

Popular repositories Loading

  1. iris-ml-predictor iris-ml-predictor Public

    A machine learning workflow for classifying iris flowers using the famous Iris dataset. It provides options for both Docker and Conda-based execution environments.

    HTML 1

  2. symbiote-lite-analyst-agent symbiote-lite-analyst-agent Public

    A lightweight human-in-the-loop AI agent that augments analyst workflows through guided reasoning, synthesis, and decision support.

    Jupyter Notebook 1

  3. omniagent-multi-agent-analytics omniagent-multi-agent-analytics Public

    OmniAgent is an AI-powered conversational data assistant with a multi-agent architecture. It automatically routes your questions to the right agent to generate insights, visualizations, aggregation…

    Python 1

  4. pipecat-voice-agent-rag-mcp pipecat-voice-agent-rag-mcp Public

    Real-time voice AI agent built with Pipecat, featuring RAG knowledge retrieval and MCP tool integration. Sub-400ms latency with Deepgram STT, Claude LLM, and Cartesia TTS.

    1

  5. agentic-rag-knowledge-ai-platform agentic-rag-knowledge-ai-platform Public

    Skein is an agentic AI platform that answers questions by weaving together structured and unstructured data, grounding every response in a knowledge graph, hybrid vector search, and a governed sema…

    Python 1

  6. tracegym-ai-agent-evaluation tracegym-ai-agent-evaluation Public

    Record, replay, evaluate, and regression-test AI agents with deterministic traces, calibrated LLM-as-a-judge evals, OpenTelemetry observability, and CI quality gates.

    Python 1