Audits, build sprints, training, and advisory for organisations
deploying AI in regulated industries. Transformation consulting with AI fluency —
an unbroken chain of judgment, design and engineering. Clarity, not hype.
01 / About
20+ years turning ambiguity into working systems
Silsila is a solo consultancy led by Faraz Fookeer — Senior BA and Product Owner
with two decades across buy-side and sell-side financial institutions, specialising
in front-office systems, collateral management, fixed income, and repo trading
technology. Currently completing an MA in AI Ethics & Society at Birkbeck,
University of London.
The name Silsila (سلسلة) means chain or unbroken transmission in Arabic —
most powerfully used in Sufi tradition to describe the lineage from teacher to student.
It reflects how I work: connecting business intent, technical reality, and human adoption
into a single, accountable chain.
02 / Engagements
How we work together
Four shaped engagements. Each has a defined duration, a named output, and a clear best-fit.
Bigger retained work emerges from clients who start here.
★ Featured
AI Training & Capability Building
For teams who need to actually use AI, not just talk about it.
Tiered programmes built from 20+ years in regulated finance and an MA in AI Ethics & Society.
A short, structured assessment of where AI creates genuine value in your operations
versus where it creates noise. Stakeholder interviews, workflow review, vendor-landscape scan.
Duration
2–3 weeks
Output
Written assessment + readout. What to build, what to buy, what to leave alone.
Best for
Leaders who suspect their AI strategy is more noise than signal.
Take one real workflow from “we want AI here” to a working prototype with guardrails.
Agent architecture, tool design, human-in-the-loop, evaluation harness.
Duration
4–6 weeks
Output
Working prototype, integration plan, risk register, evaluation results.
Best for
Teams with approval to build but unsure what to build first.
Standing senior judgement on tap. Monthly retainer for organisations who need AI strategy,
vendor evaluation, and risk thinking without making a permanent hire.
Cadence
Monthly retainer, 3-month minimum
Output
Standing time, written briefings, sounding-board access, structured deliverables on request.
Best for
Organisations that need senior AI judgement on tap but not a full-time hire.
Named projects from the last decade across tier-1 investment banks and asset managers.
Where confidentiality permits, specifics are described; where it does not, the scope is summarised.
Insight Investment
2025 · ongoing
GenAI Collateral Test Agent
Built a generative AI agent (ChatGPT-4) to run daily production parallel UAT on collateral data,
summarising movement narratives for Front Office and Middle Office stakeholders.
Delivered material efficiency gains during a critical project working group.
GenAICollateralUAT
Insight Investment
2022 · ongoing
Liquidity Manager — Repo OMS
Lead BA / Product Owner for a bond inventory and repo order management system built from the
ground up. Translated Front Office (Portfolio Managers, Investment Analysts) and Middle Office
requirements into delivered, traceable user stories enabling timely and accurate repo execution.
Repo OMSFixed IncomeProduct Ownership
Insight Investment
2022 · 2023
Collateral Monitoring — post-LDI crisis response
Lead BA on a greenfield Collateral Monitoring application, partnering with Derivatives
Operations Middle Office and managing the Northern Trust vendor relationship. Implemented
the firm's first intraday collateral data stream, materially increasing business resiliency
following the 2022 UK pension bond crisis.
CollateralVendor MgmtResilience
Northern Trust
Launched 09/2021
Bitcoin ETF Target Operating Model
Modelled the end-to-end operating framework for a Bitcoin ETF for a major client —
adopted and launched September 2021. Specified service model, custody arrangements,
NAV calculation, and regulatory touchpoints across the trade lifecycle.
Digital AssetsTOMETF
UBS
2021
Ethereum Cross-Chain DvP PoCs
Delivered three Ethereum-based proof-of-concepts — digital gold, tokenised bonds,
and tokenised debt — on time and within budget. Managed vendor oversight and
Codefi sandbox testing across the innovation programme.
BlockchainTokenisationPoC delivery
BNP Paribas CIB
2013 · 2017
Counterparty Credit Limits & Approvals Platform
Product Owner for front-to-back credit limits and rating platform delivery, managing Mumbai
development and QA teams in a hybrid SDLC. Won an internal innovation reward for a credit
request capture digital transformation built on data analysis and UI prototyping.
CCRCredit LimitsProduct Ownership
05 / Philosophy
How I work
Clarity over cleverness
The best solution is the one people actually understand and use. I optimise for adoption, not elegance in isolation.
AI fluency, not AI first
I understand what AI can and cannot do. I bring that fluency to every engagement without assuming AI is the answer to every problem.
Honest evaluation
No vendor partnerships. No commissions. No incentive to recommend anything other than what actually works for your situation.
Adab
An Arabic concept meaning proper conduct and respect. I treat every engagement — every person, every system, every constraint — with the seriousness it deserves.
06 / Projects
Selected work & builds
A mix of client-facing consulting artefacts and open-source builds from
Silsila Labs.
Silsila Labs · GitHub
RepoAgent — Funding Desk
Autonomous repo order management for fixed-income portfolios. Multi-agent system detects cash and security shortfalls (T+0 to T+2) and books the repo / reverse-repo trades that resolve them, with an adversarial Risk agent running eight checks before anything books. Python, Anthropic SDK, FastAPI, Next.js. Watch the overview
Silsila Labs · Claude Code skill · GitHub
cv-creator
A Claude Code skill that generates, reviews, and ATS-optimises CVs across five regions and seven industries. Separates regional archetypes from industry overlays so any region composes with any industry. Markdown source builds to PDF and DOCX via pandoc. View on GitHub
Silsila Labs · Onboarding kit · GitHub
Cowork Starter Pack
A no-code onboarding kit that gets non-developers productive with Claude's Cowork. Download, attach the folder, paste one prompt. Includes a printable Ask-vs-Act card, glossary, ten ranked first tasks, a connector picker, and five ready-made project workspaces. View on GitHub
Silsila Labs · Data tooling · GitHub
Dataport AI for Tableau
Turns messy CSV/Excel files into Tableau-ready data with AI-generated insights. Profiles, cleanses (every fix logged), and has Claude narrate the numbers into data stories across eight categories. Exports a .hyper extract, cleaned CSV, and HTML report. Python, Textual TUI, BYOK. View on GitHub
Live · Hugging Face
BabyGPT
Educational LLM platform built from scratch in pure NumPy. Train transformers, visualise attention, fine-tune with LoRA. No PyTorch, no TensorFlow. Try it on Hugging Face
Silsila Labs
Wraith
AI agent web extraction layer. Playwright-based browsing with TOML configs, structured extraction, workflow DAGs, and a FastAPI MCP server for agent orchestration.
Silsila Labs
Spectre
Continuous verification for AI agents. Five-dimension evaluation with LLM-as-judge plus deterministic checks, SQLite drift tracking. Open-core + managed tier from £49/mo.
Silsila Labs
TCG Trading Agent
18-module autonomous arbitrage system for Pokémon and One Piece TCG. Bloomberg-style UI, BYOK LLM router, signal engine, portfolio manager, human-in-the-loop eBay execution.
Silsila Labs
PersonaEngine
Pre-UAT stakeholder simulation. Per-persona RAG pipelines surface contradiction and concern before requirements reach development — cutting late-cycle rework.
Silsila Labs · BYOK
Agentic Scrum Team
Four specialised agents (PO, BA, Dev, QA) running an autonomous delivery loop over Linear and Slack. Built for technical PMs and BAs in regulated industries.
Silsila Labs · Live
Silsila AI Tutor
A free, scenario-driven curriculum: six tiers from how models work to governing them in regulated settings, with a scenario check after every lesson. Read it as an essay or drive it from a retro terminal mode. No sign-up. Open the tutor
Shipped · GitHub
LDI Crisis Simulator
Agentic AI LDI investment fund simulator. Autonomous agents generate market scenarios, detect insolvency risks, and execute trades across a 30-day loop. Built with n8n and Gemini 3 Pro. View on GitHub
Shipped
Rug-Pull Detector
Autonomous smart-contract auditor built with Claude Code and Sonnet 4.5. Analyses GitHub repos for rug-pull patterns in Solidity: hidden mints, honeypots, fee manipulation. Risk scoring 0–100. Python, FastAPI, React, Slither.
Shipped · GitHub
V3 Strategy Council
AI strategic advisor web application for Victoria 3 grand-strategy players. Multi-agent council surfaces optimal moves from current game state. Built with Antigravity and Gemini 3 Pro. View on GitHub
Demo
MidCap Alpha
Dual-mode React app: multi-agent Intel Mode with Research Analyst, Execution Strategist, and CIO synthesiser, plus Academy Mode equities curriculum.
07 / Contact
Let's talk
Three ways to start a conversation. The fit comes first — if I'm not the right person for what you need, I'll tell you.