AI Engineering · Security Leadership · Quantitative Systems

Kevin Blackman — AI Engineering & Security Leader

Engineering and technology leader with 20+ years building and operating production platforms, now focused on AI-augmented software engineering and real-time data infrastructure. Executive track record as a CTO and a CISO, leading multinational teams of 40+ and owning enterprise risk in regulated financial services; co-inventor of an internationally granted patent. Today I build and run KGB-ONE single-handedly — a billion-row real-time options analytics platform, developed with a self-designed multi-agent AI coding harness, and measured rather than assumed: on a fixed benchmark, the context and memory layer behind that harness is worth a 5x reduction in agent steps and an 11x reduction in cost per completed task.

Facet 01 · Skill

Security Leadership

CISO/CTO background spanning two generations of WebTrust-certified public-key infrastructure — multi-layer CAs, registration authorities, LDAP, OCSP and HSM-integrated cryptography — deployed to government, banking and industrial clients across 30+ countries under annual WebTrust audit programmes. Later, as CISO and Data Protection Officer of a CHF 14bn+ wealth and asset manager across six jurisdictions, built and ran an information-security program (ISMS policy per ISO27001 / FINMA / SOC2 / DORA / GDPR), an IT-risk portfolio, IAM/IGA oversight, board and Risk-Committee reporting, incident response and third-party vendor risk, and served as Data Protection Officer.

Applied Security Engineering

Practices what he preaches on his own platforms: isolated read-only public services, least-privilege database roles, hardened containers (non-root, read-only rootfs), TLS/CSP discipline, firewall and fail2ban operations, and recurring security audits of his own internet-facing surfaces.

Governance, Risk & Compliance

ISO27001, FINMA, DORA, SOC2, ISAE3402 and GDPR programs; 24×7 SOC/SIEM delivery ($1M+ programs); EPP / MDM / 2FA transitions; board and Risk-Committee reporting; Data Protection Officer.

PKI, WebTrust & Cryptography

Earlier career · 2001–2015

Two generations of WebTrust-certified public-key infrastructure — multi-layer CAs, registration authorities, LDAP, and OCSP — deployed to government, banking, and industrial clients across 30+ countries. HSM-integrated custom cryptography, smart-card platforms, and a granted patent for digital authentication of valuable goods. Secure data centres built to WebTrust and DoD Level 5 specifications; annual WebTrust audit programmes; EU EEMA PKI interoperability standards work.

  • PKI
  • WebTrust
  • Cryptography
  • HSM
  • OCSP
  • X.509

Digital Identity, Social Networks & Blockchain

Earlier career · 2001–2015

Secure credential and password-management platforms across web, desktop, and mobile — X.509 digital IDs, encrypted local stores, and biometric integration. Consumer social-network systems for major sports franchises with Facebook/Twitter APIs, content management, and fan push alerts. Smart-card authentication architectures; commercial certificate and subscription transaction portals. Cryptocurrency and blockchain identity infrastructure for digital-asset provenance and secure transactions.

  • Social networks
  • Blockchain
  • Cryptocurrency
  • iOS
  • Android
  • Biometrics
  • Smartcards

Security writing

Facet 02 · Skill

AI & LLM Systems

I treat agent performance as an engineering problem rather than a matter of taste. The harness is the hard part, not the model: heterogeneous vendor agent CLIs coordinating with no central orchestrator, through a shared verified context and a task queue, with per-task git-worktree isolation, a correctness gate before any work is accepted, scored task lineage under a beat-the-incumbent commit rule, and cost-disciplined routing across 30+ models and 8 providers. Agents carry lessons between runs, so a known-failed approach is not retried.

Multi-Agent Orchestration Stack

Multi-agent orchestration over a knowledge base of thousands of embedded research chunks with structured tool use, RAG and knowledge-graph pipelines, and tiered dispatch across multiple LLM providers. LLM-based document and report generation is gated by adversarial QA before publication. The stack learns from daily write-ups and past sessions via pgvector memory.

  • pgvector
  • Neo4j
  • PostgreSQL
  • vLLM
  • Ollama
  • GraphRAG
  • Multi-agent

Agent Memory & Evaluation

Agents that learn across tasks — each run distilled into a lesson and replayed to related work, so failures are not repeated. Evaluated the way infrastructure should be: fixed benchmarks, ablation studies and cost-per-completed-task measurement, with pgvector semantic search and GraphRAG retrieval underneath.

  • Benchmarks
  • Ablation studies
  • Cost routing
  • pgvector
  • Lesson replay

AI writing

Facet 03 · Skill

Quant & Systems Engineering

Real-time market-microstructure infrastructure and a strategy-research engine, built and operated solo: 43 services in production, a 275 GB TimescaleDB holding 1.06 billion rows, and 421 reproducible backtest runs with full metadata. Dockerised microservices, real-time WebSocket ingestion, Redis distribution and a columnar-cache layer for bulk-replay research with walk-forward validation and prop-firm survival modeling — plus full infrastructure ownership: networking, firewalling, secure remote access, observability and automated alerting.

KGB-ONE Options Intelligence

A real-time SPX options analytics and automated-reporting platform: per-minute dealer-exposure snapshots, filtered net-premium drift, regime detection, and gated end-of-day / pre-open / intraday report generation — surfaced through a React charting portal, with verification gates on every published figure.

  • Python
  • TimescaleDB
  • Redis Streams
  • React
  • Docker

Research & Backtest Engine

Backtesting, a strategy registry and signal research (dealer-positioning, regime and ML families) with prop-firm survival simulation and reproducible, cached result sets for comparable evaluation.

  • Backtrader
  • pandas
  • NumPy
  • Machine Learning

Real-Time Data Engineering

DXLink / WebSocket ingestion, TimescaleDB hypertables, Redis Streams event bus, Dockerised microservices on Linux with systemd; FastAPI backends and React + Vite + Tailwind front ends with Plotly interactive charting.

Facet 04 · Skill

Trading & Markets

Defined-risk options strategies on SPX and index products, driven by dealer positioning and regime — carrying strong Calmar and Sortino ratios across forward testing and live management. The options analytics behind this work — full chain, historical and live, with millisecond-accurate real-time flow — come from a commercial provider. Live trading integration spans broker and execution APIs across Tastytrade, Schwab (thinkorswim), MooMoo (Futu OpenD API) and cTrader (prop-firm FIX / Open API + Protobuf).

Facet 05 · Skill

Career & Credentials

Swiss national (Geneva) · bilingual English / French. Executive MBA, Management of Technology, EPFL Lausanne (2010 — Prize for Best Grades, HEC/EPFL, and Best Business Plan award). BSc Electrical & Computer Engineering, University of the West Indies (1995, 1st in Major). PMP (2025), CISSP, ISSAP, ISSMP, CSSLP and Python developer certification. CEIBS Shanghai "Doing Business in China" certificate (2009). ISC2 exam-item review volunteer since 2005.

Core competencies

  • AI engineering & agentic systems — multi-agent orchestration with no central orchestrator (DeLM), correctness-gated verification, git-worktree task isolation, scored task lineage with a beat-the-incumbent commit rule, cost-disciplined routing across 30+ models and 8 providers
  • Agent memory & evaluation — lessons distilled per run and replayed to related work; fixed benchmarks, ablation studies and cost-per-completed-task measurement; pgvector semantic search and GraphRAG retrieval
  • Engineering leadership — team building (40+ FTEs), Agile delivery across product, infrastructure and mobile platforms, vendor management, multi-site global operations, 0-to-1 product ownership
  • Systematic options trading — defined-risk multi-DTE strategies; risk-adjusted performance (Calmar, Sortino) across forward testing and live management
  • Real-time market-data engineering — DXLink / WebSocket ingestion, TimescaleDB hypertables, Redis Streams event bus, Dockerised microservices on Linux with systemd
  • Options analytics — GEX / DEX / VEX, charm, dealer positioning, max-pain, skew & term structure
  • Quantitative research & ML — Backtrader walk-forward validation, prop-firm survival modeling, scikit-learn / XGBoost signal work; Parquet-derived cache layers for bulk pandas replay (Postgres as source of truth, columnar extracts where SQL would be the wrong tool)
  • Live trading integration — broker and execution APIs across Tastytrade, Schwab (thinkorswim), MooMoo (Futu OpenD API), and cTrader (prop-firm FIX / Open API + Protobuf); DXLink market-data feeds
  • Full-stack web engineering — FastAPI backends, React + Vite + Tailwind front ends, Plotly interactive charting and research portals
  • Applied AI & agent systems — multi-agent orchestration, RAG over pgvector and Neo4j, tiered LLM dispatch with rate-limit metering, structured tool use, self-hosted inference (vLLM / Ollama)

Career timeline

Earlier: UN/UNCTAD technical consultant (PKI design, Secure Electronic Commerce Authenticated Link project, 1997); Project Manager for the Science, Technology & Innovation Park phase I at TIDCO, Trinidad & Tobago (USD 30M+ budget, reported to ministerial cabinet). Volunteering: JCI Seniors Geneva, president (2+ years); MassChallenge Switzerland, program coordinator.

Tech stack

  • Python
  • FastAPI
  • Pandas
  • Parquet
  • scikit-learn
  • XGBoost
  • Backtrader
  • PostgreSQL
  • TimescaleDB
  • Redis
  • pgvector
  • Neo4j
  • Docker
  • systemd
  • Linux
  • WebSockets
  • Protobuf
  • React
  • Vite
  • Tailwind CSS
  • Plotly
  • DXLink
  • cTrader
  • Tastytrade
  • Schwab
  • Futu API
  • vLLM
  • Ollama
  • RAG
  • LLM dispatch
  • Multi-agent

Availability

Open to developer evaluation programs, compute-grant collaborations, technical case studies, security and technical roles, consulting and speaker engagements.

Daily intelligence

Daily Market Briefings

Now published at blackmantrading.com

The daily end-of-day market briefings — dealer-positioning, regime, flow and the levels that matter, fronted by $RAVOLM, a regime-aware animated volume visualisation — now live at blackmantrading.com.

Get in touch

Let's talk

For interesting opportunities, ventures, evaluation programs, compute grants, technical case studies, or quantitative work — the fastest way to reach me is email.