Shipping ML systems at the edge of finance, risk, and retrieval.
📍 New Jersey, USA · 🎯 Open to senior AI/ML roles
| 5+ yrs | 35% | 25% | 30% | 40% |
|---|---|---|---|---|
| Production AI/ML | RAG retrieval accuracy |
Risk-model accuracy |
Faster model training |
Faster data queries |
I'm an AI/ML engineer with 5+ years of production experience building ML and Generative AI systems. Currently at Goldman Sachs, building financial research and risk-intelligence platforms — RAG pipelines, agentic LLM workflows, and models for risk prediction, anomaly detection, and fraud.
Before Goldman Sachs, I spent two and a half years at Cognizant as a Data Scientist — building real-time market models, reinforcement-learning execution strategies, anomaly-detection frameworks, and NLP sentiment pipelines over large-scale financial data. M.S. in Computer Science from SUNY Albany.
My work sits where modeling meets engineering: latency budgets, evaluation that survives production, and pipelines that don't fall over when the data changes underneath them. Offline accuracy is the start of the job — shipping is the rest.
gantt
title Industry experience
dateFormat YYYY-MM
axisFormat %Y
section Roles
Data Scientist · Cognizant :done, c1, 2020-02, 2022-08
AI / ML Engineer · Goldman Sachs (current) :active, m1, 2023-08, 2026-06
A multi-agent retrieval-augmented system where specialized agents handle routing, retrieval, query reformulation, fact-checking, and safety. Built with LangChain + Streamlit + Groq, hybrid file/URL knowledge sources, dynamic routing and self-correction.
flowchart LR
Q([User query]) --> R{{Router agent}}
R -- "files / URLs" --> Ret[Retriever]
R -- "fresh facts" --> Web[Web search agent]
Ret --> Rf[Query reformulator]
Web --> Rf
Rf --> Fc[Fact-checker agent]
Fc --> Sc[Safety-checker agent]
Sc --> Out([Grounded response])
classDef agent fill:#1f6feb22,stroke:#1f6feb,stroke-width:1.5px,color:#1f6feb
classDef io fill:#d4ff3a22,stroke:#8ba526,color:#3d3a35
class R,Ret,Web,Rf,Fc,Sc agent
class Q,Out io
➜ github.com/ms1104n-max/multi-agentic-rag
| Repo | What it actually does |
|---|---|
multi-agentic-rag |
Multi-agent RAG with router · retriever · reformulator · web-search · fact-checker · safety agents (Streamlit + LangChain + Groq) |
rag-chatbot |
Production-aware RAG chatbot · CI/CD · M1 / NVIDIA llama.cpp · explicit cost / latency / hallucination tradeoffs |
rag-from-scratch |
RAG without framework abstractions — embeddings, local vector DB, retrieval, re-ranking, query rewriting from first principles |
langchain-rag-document-understanding |
LangChain + FAISS + SentenceTransformer pipeline for grounded document Q&A · Jupyter walkthrough |
mlops-app |
IaC reference stack — Terraform on GCP · BigQuery · GH Actions · Docker · Prefect · dbt · MLflow · FastAPI |
ai_soc— hands-on study of Srinivas et al., AI-Augmented SOC: A Survey of LLMs and Agents for Security Automation (Informatics, 2025). Implementation walkthrough used to learn multi-agent security architectures and ML-driven threat detection. Original implementation by Abdul Bari (CSU San Bernardino) — attribution preserved per Apache 2.0.
- 35% higher retrieval accuracy · production RAG pipelines for financial research & knowledge discovery — Goldman Sachs
- 25% accuracy lift · risk-prediction, anomaly-detection, and fraud models — Goldman Sachs
- 30% faster · model training and large-scale data processing with JAX — Goldman Sachs
- 40% faster · query response on a 100 TB+ Azure Synapse warehouse — Cognizant
- 30% fewer false positives · real-time anomaly detection (Kafka + PySpark), SEC / FINRA aligned — Cognizant
Full case studies & architecture detail at sainikhil.com →
mindmap
root((Stack))
AI and ML
Python
Scikit-learn
TensorFlow
Keras
JAX
Hugging Face
Generative AI
LangChain
OpenAI
Azure OpenAI
RAG
Embeddings
LoRA PEFT
Data
PySpark
Pandas
NumPy
Kafka
Snowflake
Postgres
MongoDB
Vector DBs
FAISS
Pinecone
ChromaDB
Cloud
AWS
AWS Lambda
Vertex AI
Azure Synapse
MLOps
Docker
Kubernetes
MLflow
Jenkins
FastAPI
CICD
pie showData
title Focus distribution
"Generative AI · RAG · Agents" : 35
"Financial ML" : 30
"MLOps & Backend" : 20
"Data Engineering · Analytics" : 15
- 📍 Open to senior AI/ML and Generative-AI roles
- 🛠️ Shipping financial research & risk-intelligence systems at Goldman Sachs
- 📫 Reach me at smatt11041509@gmail.com
Production-first ML — offline accuracy is the start of the job; shipping is the rest.