Senior Data Engineer -> Quant Developer

Building Streaming Market Data and AI-Assisted Trading Systems

I design low-latency data pipelines, ClickHouse analytics layers, deterministic trading strategies, and LLM-assisted signal evaluation for stock market research and execution workflows.

View QuantStream Resume
Shashidhar Reddy

About

Data engineering foundation with a quant systems focus.

I am a senior data engineer experienced in Python, SQL, Spark, AWS, Airflow, NiFi, streaming pipelines, and analytical data platforms. I am now applying that engineering background to quant developer and quant data engineering opportunities.

My current focus is QuantStream: a modular research and trading platform that turns real-time ticks, order flow, option metrics, and footprint candles into explainable strategy signals, AI scores, trade logs, and replay-ready datasets.

Realtime Market Data
ClickHouse Analytics
LLM Signal Scoring

Featured Project

QuantStream - AI-Assisted Order Flow Trading Platform

Institutional-style automated options research system for stock market, indices, and F&O instruments. The platform ingests live market data, builds footprint and candle layers, detects order-flow imbalances, confirms breakouts, and evaluates candidate trades with Anthropic Claude or local open-source LLM models.

01

WebSocket Layer

Broker APIs and order-flow data providers for ticks, futures, order flow, OI, IV, bid/ask, and footprint data.

02

Streaming Layer

Redpanda/Kafka-compatible ingestion with compact analytics payloads, compression, and source-specific topics.

03

Storage Layer

ClickHouse raw tables, materialized views, 30s/1m/5m rollups, footprint levels, retention, and replay queries.

04

Feature Layer

Deterministic features: cumulative delta, stacked imbalance, absorption, VWAP, SuperTrend, RSI, IV, and OI context.

Reusable Quant Architecture

Foundation layer for any strategy

Every strategy starts from the same explainable pipeline: capture market events, store replayable data, filter deterministic signals, evaluate quality with AI, and route approved setups through risk-aware execution.

WebSockets Ticks, order flow, options, F&O
live stream
Storage Layer Redpanda, ClickHouse, rollups, replay
raw candles
Analyze Layer Feature engineering, filters, signal rules
delta VWAP
AI Layer Evaluation, scoring, confidence, reasoning
LLM JSON
Execute Layer Risk checks, dry-run logs, broker routing
risk orders

Strategy Layer

Deterministic candidate generation from 5-minute footprint candles and 1-minute confirmation closes. Strategies stay explainable and testable.

  • Order-flow imbalance detection
  • Opening Range Breakout research
  • Volume shocker scans and stock F&O backtests

AI Evaluation Layer

The LLM is a trade-quality evaluator, not a blind predictor. It consumes engineered features and returns structured JSON for storage and analysis.

  • Anthropic Claude Haiku for low-latency scoring
  • Ollama/open-source model benchmarks
  • Prompt compression, JSON validation, token and latency logging

Risk + Execution Layer

Execution is isolated from strategy logic. Current mode focuses on dry-run trade logging, trailing stop ladders, and outcome tracking before live gates.

  • Broker execution integration
  • Trade setup rows and outcome tracker
  • Replay-ready logs for win-rate, drawdown, and expectancy analysis
5m + 1m Footprint signal and confirmation workflow
Redpanda Kafka-compatible event streaming
ClickHouse Fast rollups, replay, and AI score storage
LLM JSON Explainable TAKE, SKIP, WAIT decisions

Skill Set

Quant data engineering, AI, and production systems.

Quant Finance

Order flow, footprint candles, cumulative delta, VWAP, SuperTrend, RSI, OI/IV context, volatility regimes, options analytics.

Data Engineering

Python, SQL, Spark, AWS, Airflow, NiFi, Redpanda/Kafka, ClickHouse, Docker, Linux, systemd, schema design.

AI + LLMs

Anthropic Claude, Ollama, open-source LLM testing, structured prompts, compact codecs, JSON contracts, model benchmarking.

Research + Risk

Backtesting, replay pipelines, signal logs, confidence scoring, dry-run execution, trailing stops, outcome tracking, risk gates.

Project Portfolio

Data platforms, fintech systems, and quant research.

QuantStream
Quant Finance

QuantStream

Streaming market data, order-flow strategies, ClickHouse features, and AI-assisted signal scoring.
Fintech Data Specialist

Enable Benefits

Low-latency, highly compliant data pipelines for secure document intake, historical data ingestion, CRA rule auditing, and client lifecycle analytics.
TUI Group
Data Analytics

TUI Group UK

Big data engineering and analytics for one of the world's leading travel companies.
JUMO
Data Systems

JUMO

Senior data engineering for mobile financial services, lending, savings, and EDW workflows.
Old Mutual South Africa
AWS Data Engineering

Old Mutual SA

Cloud data pipelines, production data movement, and analytics engineering.

Get In Touch

I am interested in quant developer, quant data engineer, and real-time data platform roles.

GitHub
LinkedIn
Email
PHP Hits Count
PHP Hits Count