Aviral Anand
Mishra
I own the full stack of a live derivatives trading operation — from decoding raw exchange multicast in C++ at sub-microsecond latency, through Greeks pricing and volatility research, to a unified multi-broker OMS running real capital.
Mumbai · dSₜ = μSₜdt + σSₜdWₜ · microstructure × execution
$ ./feed_decoder --segment FO --transport multicast
[09:15:00.000001] UDP feed sync · NSE FO · p99 decode 0.9μs
[09:15:00.000114] L2 book rebuilt · depth 5 · seqlock shm publish
[09:15:00.104520] IV solve (Newton–Raphson) σ = 13.42% · 6 iters
[09:15:00.104615] Δ +0.42 Γ 0.031 ν 11.2 Θ −8.4
[09:15:00.201773] ORDER NIFTY 24800 CE SELL 50 → FILLED @ 142.30 · slip 0.4 bps
[09:15:00.201801] fanout 500 clients · 32ms · 0 rejects
635×
Order-Fanout Latency Reduction
4.2 TiB
Tick-Data Warehouse (ClickHouse)
6+
Broker APIs Behind One OMS
20+
Strategies Built & Backtested
01 — Who I Am
About Me
I'm a quant researcher-engineer who owns the full stack of a live derivatives trading operation — from decoding raw NSE exchange multicast feeds in C++ at sub-microsecond latency, through signal generation and Black-Scholes/Greeks pricing, to a unified multi-broker order management layer running real capital across 6+ brokers. I move comfortably between quant research (statistical arbitrage, ML-based option strategies, IV-surface modeling), low-level systems engineering (lock-free shared memory, UDP protocol parsing, ClickHouse at multi-TB scale), and production backend work — maintaining 50+ repositories spanning strategies, OMS platforms, data infrastructure, and tooling.
Trading Systems
Production-grade OMS, execution engines, multi-broker order routing, and strategy deployment pipelines for live systematic trading.
Market Data Infrastructure
Raw exchange multicast decoding, L2 orderbook reconstruction, and low-latency streaming distribution across NSE, BSE, and MCX.
Quantitative Research
Statistical arbitrage, volatility-surface modeling, regime classification, Monte Carlo simulation, and derivatives strategy backtesting.
Risk & Monitoring
SPAN margin computation, real-time position reconciliation, execution anomaly detection, and centralized logging infrastructure.
02 — Skills & Credentials
The Quant Arsenal
Δ
Delta
directional exposure & hedging
Γ
Gamma
convexity & rebalance risk
ν
Vega
volatility surface exposure
Θ
Theta
decay harvesting on 0DTE
Languages
Quant Methods
Low-Latency Systems
Trading Systems
Data & Infrastructure
Broker Integrations
Certifications
03 — Measured, Not Claimed
Engineering Highlights
Sub-Microsecond Feed Decoding
C++ decoder for NSE direct multicast UDP feeds — 10+ transaction codes across CM and FO segments — benchmarked at sub-microsecond p99 over 500,000 iterations, publishing to lock-free shared memory and Redis simultaneously.
635× Order-Fanout Speedup
Rebuilt the signal→order hot path of a multi-client OMS with async parallelism and fire-and-forget persistence; a 500-client order placement benchmark dropped from 20.3 seconds to 32 milliseconds.
IV Solver in Native SQL
Implied-volatility solver written directly in ClickHouse SQL — a 20-iteration Newton–Raphson using arrayFold and erfc — cross-validated to ~1e-8 against py_vollib and scipy.
Multi-TB Tick Analytics Platform
Self-hosted ClickHouse cluster with a 4.2 TiB tick table growing ~7 GiB/day across 600+ daily partitions, plus a 140 GiB enriched-options analytics table with memory-bounded parameterized views.
Unified C++ Order Orchestrator
C++20 daemon that validates, slices by exchange freeze quantity, routes, and tracks fills across six brokers through one adapter layer — any strategy, in any language, places orders via a single gRPC interface.
Margin Engine From Scratch
Production margin service parsing daily NSE SPAN risk files across all product families — SPAN, exposure, and premium margin — later extended to MCX commodities. 17/17 tests passing on first delivery.
04 — Strategy Classes
Quant Research
Δ
Delta-Neutral Structures
Multi-leg, delta-targeted option structures with continuous Greeks monitoring and systematic position management — a strategy class deployed live across multiple accounts and brokers.
Γ
Short-Volatility Structures
Defined-risk premium-selling structures on index derivatives — systematic short-volatility positioning with mark-to-market risk limits and time-based exits.
ρ
Statistical Arbitrage
Cointegration-driven pairs and mean-reversion research on Indian equities — staged from research notebooks through paper trading to real-time engines.
σ
Volatility Surfaces
SVI volatility-surface fitting, IV-percentile research, and index dispersion strategies — implied vol treated as a first-class research object, solved and stored at tick scale.
𝔼
ML Signal Generation
Gradient-boosting models for option-selling signals, Markov-chain regime models on intraday bars, and walk-forward validation with strict no-lookahead discipline.
f*
Sizing & Risk
Kelly-criterion position sizing, Monte Carlo robustness testing, and investor-grade reporting standardized on CAGR, Sharpe, Sortino, Calmar, and drawdown before any number ships.
Every strategy graduates through a staged pipeline — research notebook → deterministic no-lookahead backtest → paper trading → live deployment — with walk-forward validation and investor-grade reporting (CAGR, Sharpe, Sortino, Calmar, drawdown) before any capital is touched.
05 — Infrastructure I've Built
Trading Systems
Multi-Broker Execution Engine
Automated trade execution framework integrating multiple broker APIs (Zerodha, Kotak, XTS, Motilal Oswal) with smart order routing, real-time position tracking, and PnL & exposure monitoring.
L2 Orderbook Engine
Real-time L2 orderbook reconstruction from NSE UDP multicast feeds. Binary protocol parsing, top-5 depth management, lock-free shared-memory publishing, tick normalization & streaming.
Market Data Distributor
Scalable pub/sub architecture distributing thousands of normalized tick streams to multiple strategies simultaneously. Multi-consumer design with replay capability for research.
Options Analytics Engine
Real-time Greeks computation & implied volatility modeling. Delta, Gamma, Vega calculations with strike filtering, exposure mapping, and multi-leg payoff evaluation.
Strategy Backtesting Framework
Modular engine for equity & derivatives strategy simulation over multi-year minute data. Transaction cost modeling, multi-leg support, M2M integration & Sharpe/drawdown risk metrics.
Execution Analytics Module
Trade-level and strategy-level performance evaluation. Slippage analysis, risk-adjusted return tracking, and Sharpe & drawdown computation.
ClickHouse Backtesting Engine
Parameter-driven backtesting engine on ClickHouse supporting multi-leg derivatives strategies (iron condors, premium structures). Auto-generates Sharpe, drawdown, P&L, and trade-level statistics.
Tick-Level Market Replay
Historical microstructure replay engine reconstructing tick-by-tick market data for analysis of execution timing, price behavior, and strategy reactions to real order flow.
Risk & Monitoring Systems
Real-time monitoring for margin utilization, abnormal fills, and execution anomalies. Monte Carlo simulation for strategy robustness and centralized structured logging.
06 — Papers & Books
Published Work
Amity Journal of
Computational Sciences
VOL 5 · ISSUE 2 · 2021
Quantum
Computing
|ψ⟩ = α|0⟩ + β|1⟩
Aviral Anand Mishra
Quantum Computing: The Next Revolution in Computing
Aviral Mishra · Dr. Ranjana Rajnish — Amity University, Lucknow · ISSN 2456-6616
Published in the Amity Journal of Computational Sciences (Vol 5, Issue 2, 2021), this study explores quantum computing as the next paradigm shift in computation — how superposition and entanglement let a qubit hold far more than a classical bit's 0 or 1. It surveys real-world application areas — AI & machine learning, computational chemistry, drug discovery, quantum cryptography, financial modelling, logistics, and weather forecasting — alongside the field's hard open problems: decoherence, error correction, and qubit quality.
Written during my undergraduate years at Amity University Lucknow — the same first-principles curiosity that now drives my work on volatility surfaces and low-latency systems, one layer deeper: |ψ⟩ = α|0⟩ + β|1⟩.
07 — What's Next
Current Focus
Scaling multi-broker OMS for higher throughput and more accounts
Advancing execution quality with price-chasing and best-entry algorithms
Expanding the systematic derivatives strategy portfolio
Deepening C++ for performance-critical trading components
σ(opportunity) > 0
Get in Touch
Open to interesting quant, trading systems, and infrastructure projects. Feel free to reach out!