Senior / Staff Software Engineer in London with 20+ years of experience building trading, pricing, risk, and post-trade systems for investment banks and financial technology firms.

Currently working at Citi via JUXT (part of Grid Dynamics), focused on distributed cross-asset risk orchestration and intraday/EOD risk processing.

LinkedIn | GitHub

• Front-office pricing, risk, and trade-lifecycle platforms
• High-throughput and event-driven JVM systems
• Low-latency, concurrent, and memory-sensitive applications
• Kafka, FIX, REST, and gRPC integration
• AI-assisted engineering workflows for implementation, testing, and review

• Citi via JUXT — cross-asset risk orchestration and intraday/EOD risk processing (March 2026 - present)
• Morgan Stanley — front-office pricing and risk for CDS Index Options and Structured Credit (September 2023 - February 2026)
• CMC Markets — low-latency options pricing and FIX connectivity using Chronicle Map off-heap storage (March 2023 - September 2023)
• Blockchain.com — institutional prime brokerage and treasury automation across Coinbase, Kraken, Binance, and Bitfinex (April 2021 - January 2023)
• Goldman Sachs and Credit Suisse — earlier engagements across equities booking, securities lending, market risk, and reference-data platforms

An end-to-end slice of a front-office platform — live market data, a matching engine, post-trade booking, risk, and one presentation layer over all of it. Five separately deployed, separately tested systems. trading-system and trading-desk depend on orderbook and risk-engine as versioned libraries rather than duplicating them — bounded contexts composed, not merged into one codebase.

▶ Explore it live: https://desk.damianhoward.com

How it fits together: a data pipeline and a presentation layer. Data: market-data anchors orderbook's book to a real price; every match publishes a fill to Kafka; trading-system consumes that stream, books the position, and reprices it by calling risk-engine as a library. Presentation: trading-desk reverse-proxies orderbook's live book and trading-system's dashboard as tabs in one shell — trade on the book, watch the position reprice beside it. risk-engine's interactive pricer runs standalone at https://risk.damianhoward.com.

• market-data — pulls real quotes from Yahoo Finance and serves the last-good snapshot, so a transient provider failure never blanks the live book it feeds.
• orderbook — a thread-safe limit order book with three interchangeable concurrency strategies, JMH-benchmarked to the nanosecond; the LMAX Disruptor implementation beats a read/write lock by roughly 6× under contention. Seeds itself from market-data's real quotes and publishes every fill to Kafka.
• risk-engine — Black-Scholes pricing and Greeks hand-written in Kotlin, cross-validated against OpenGamma Strata as an independent oracle. Runs both as its own live pricer and as the library trading-system calls on every fill.
• trading-system — consumes orderbook's fill stream off Kafka, books net positions into an Oracle Autonomous Database, reprices through risk-engine, and pushes live positions, VaR, and PnL to a dashboard. Poison messages route to a dead-letter topic after bounded retries.
• trading-desk — the live link above: a reverse-proxy gateway over the live order book and trading dashboard.

• portfolio-manager — authenticated exchange clients for Binance and Bitfinex, with venue-local HMAC signing and a withdrawal workflow that's dry-run by default and requires explicit confirmation before it touches money.
• stocks-analysis-us — a six-stage, event-driven pipeline that builds a ranked US equity universe from public fundamentals and exports it to Excel.

Smaller repos: kafka-streams-patterns (four Kafka Streams topologies), sudoku-dancing-links (Knuth's Dancing Links vs. naive backtracking), kotlin-blockchain (proof-of-work and UTXO mechanics), and bank-csv-to-qif (a CSV-to-QIF converter for legacy finance tools).

Contributed to Meridian, JUXT's equity-derivatives post-trade risk accelerator. Meridian supports valuation, Greeks, scenario analysis, and continuously updating risk on a bitemporal datastore.

My work covered the ticking-risk engine, scenario-analysis workflow, and resilient recovery of long-running valuation tasks across Kotlin, Python/QuantLib, and TypeScript. I used Claude Code as part of an agentic engineering workflow spanning implementation, testing, and review — the same workflow behind the repositories above.

I also contributed to a privately developed AI-assistant platform, delivering a cross-platform notifications service for alerting, validated response capture, and scoped delivery across distributed services.

• Languages: Java, Kotlin, Scala, Python, TypeScript
• Trading and integration: FIX, Kafka, REST, gRPC
• Platforms: OpenShift, AWS, GCP, Docker
• Domains: pricing, risk, trade lifecycle, post-trade, prime brokerage, treasury automation

For professional enquiries, contact me through LinkedIn.