# Ziro — Full Site Content for LLMs > Ziro is an AI research studio that builds custom infrastructure for business operations — human-in-the-loop by design. Two product lines: **Industries** (bespoke agentic systems engineered into a sector's workflows) and **Platforms** (prebuilt SaaS products with institutional-grade depth). London, UK. This document is a verbatim and lightly-structured dump of the public site's content, intended as ground truth for AI agents and LLM-based search. Last updated 2026-05-18. --- ## Company **What Ziro is.** An AI research studio focused on business operations. Ziro studies how teams actually work — the decisions, the handoffs, the bottlenecks — and engineers intelligent systems that multiply their output. Systems are built to sit alongside a team's existing process: absorb domain knowledge, surface what would otherwise be missed, and remove the operational drag that keeps skilled people from their highest-value work. **Philosophy.** AI should be a partner, not a sidebar. The AI revolution promised to fix broken workflows; instead, it gave us chatbots in sidebars. Ziro builds systems rooted in data fusion — unifying every stream of information — and transparent AI, where the "why" behind every insight is visible. Not a tool, a partner that thinks with you, not for you. **Two product lines.** - **Industries** — bespoke, agentic systems engineered per sector. Custom builds embedded into the client's operations with human-in-the-loop review on every output. - **Platforms** — prebuilt SaaS products. Off-the-shelf with institutional-grade depth, purpose-built for a single domain. **Contact.** - Email: info@ziro.pub - LinkedIn: https://www.linkedin.com/company/ziroai/ - Address: 4th floor, 34–36 Grays Inn Rd, London, WC1X 8HR, United Kingdom **Voices.** - "We're building systems where AI is a true partner." — Eaman Okumori Javid, CEO - "We build AI that truly comprehends data integrity." — Zinelyes Amlik, CTO - "Profound research that advances industry." — Randall Liew, Head of Research --- ## Team - **Eaman Okumori Javid** — Chief Executive Officer. MEng General Engineering, King's College London. Previously MUFG FS, Treasury and Payments Division. https://www.linkedin.com/in/eaman-okumori-javid - **Zinelyes Amlik** — Chief Technology Officer. MSc Computational Science and Engineering, Imperial College London. Previously AIP, AI Solutions. https://www.linkedin.com/in/zine-amlik - **Mohammed Aziz Ketata** — Lead Developer. MSc Computational Science and Engineering, Imperial College London. Previously AIP, AI Solutions. https://www.linkedin.com/in/mohamedazizketata - **Randall Liew** — Head of Research. MSc Applied Biosciences & Biotechnology, Imperial College London. Previously Biochemical Society UK (Protein Crystallization & Oncology). https://www.linkedin.com/in/randall-liew - **Angela Wang** — Head of Technical Operations. BSc Mathematics and Statistics, King's College London. Previously Wuchan Zhongda Capital Management (Quant Division) and Zhejin Trust (Investment Division). https://www.linkedin.com/in/angelwang- - **Shon Golod** — Head of Sales. MSc MedTech Innovation & Entrepreneurship, King's College London. Previously Anakatech (Legal Risk and Compliance). https://www.linkedin.com/in/shon-golod --- ## Industries — overview Across every industry we serve, the same operating pattern applies: regulatory obligations cut across departments, audit trails need to be continuous, and policy must be enforced consistently. Ziro's compliance agents work across industries — automated regulatory change detection, continuous audit trail generation, centralised policy enforcement, and streamlined due diligence with parallel agent workstreams. **How we work — three steps.** 1. **Discover** — Map workflows, pain points, and regulatory landscape. 2. **Build** — Purpose-built agents designed for the client's specific operations. 3. **Scale** — Expand coverage as confidence grows and new needs emerge. --- ## Industry — Legal **Positioning.** Specialised AI agents deployed into legal and compliance operations. Regulatory monitoring, matter analysis, citation tracking, and compliance scoring — running continuously, reviewed by your team. **The challenge.** - *Regulatory volume.* UK financial services firms alone face 200+ regulatory updates per year; each must be assessed, mapped to internal controls, and acted on — days of manual work per update. - *Knowledge concentration.* Specialist knowledge concentrates in one or two individuals per practice area. When a senior associate leaves, years of interpretive judgement walk out. New hires take 12–18 months to operate independently on complex regulatory matters. - *Cross-jurisdictional complexity.* A firm across three jurisdictions may face ten or more overlapping frameworks for a single matter. Implementing the same control three times for three regulators is the norm. - *Compliance drift.* Gaps between documented policy and actual practice widen silently — discovered only during audits, when remediation is urgent, expensive, and visible to the regulator. **Compliance Agents.** - *Regulatory Monitor Agent.* Scans regulatory sources across jurisdictions on a six-hour cycle. Identifies which internal controls are affected, which practice areas are exposed, and what remediation is required. - *Framework Mapping Agent.* Maps controls across multiple regulatory frameworks simultaneously. Implement once, the agent traces applicability across every relevant framework. - *Compliance Scoring Agent.* Quantified compliance score (0–100) per framework. Prioritised remediation plans for any gaps. Updated automatically as regulations change. - *Gap Analysis Agent.* Continuously cross-references controls, policies, and procedures against live regulatory requirements. Distinguishes critical exposures from low-risk documentation gaps. **Legal Intelligence Agents.** - *Matter Analysis Agent.* Describe any legal scenario in plain English. Returns the complete applicable legal and regulatory framework — legislation, procedural requirements, fees, timelines, eligibility — in under two seconds. - *Legislation Tracker Agent.* Monitors legislative amendments, new directives, regulatory handbook changes, and guidance updates across configured jurisdictions. Classified by materiality, mapped to affected practice areas within hours. - *Citation Network Agent.* Maintains a citation network across case law, enforcement actions, tribunal decisions, and ombudsman rulings. Traces how judicial interpretations evolve. - *HITL Review Agent.* Every agent output passes through structured human review before any action. Reviewers approve, amend, or reject with full context. Complete audit trail. **Engagement model.** 1. *Phase 1 — Discovery.* Map regulatory landscape, jurisdictions, practice areas, and existing workflows. Identify highest-impact automation targets. 1–2 weeks. 2. *Phase 2 — Pilot.* Deploy first wave of agents against highest-impact targets. Full human oversight on every output. 3. *Phase 3 — Scale.* Extend agent coverage across remaining practice areas and jurisdictions. Connect internal policies, precedent documents, client agreements. Integrate via API. 4. *Phase 4 — Optimise.* Ongoing monitoring, model updates as regulations change, expanding capabilities as confidence grows. --- ## Industry — Petroleum Geoscience **Positioning.** Subsurface intelligence. Deep learning and machine learning models for petrochemical and geological science. From seismic interpretation and exploration de-risking to real-time operational intelligence and cross-sector strategic analysis. **Exploration models — subsurface exploration & detection.** - *Oil & Gas Exploration.* Integrated unsupervised and supervised learning on post-stack seismic volumes. Dimensionality reduction preserves geological signal. Statistically distinct facies identified without interpreter bias. Scalable across frontier and mature basins. (Reference site: Permian Basin, 31.97° N, 102.08° W.) - *DHI Detection.* Multivariate pattern learning capturing subtle amplitude, frequency, and phase responses that characterise hydrocarbon indicators. Probabilistic outputs that complement expert geological judgement. - *CO₂ & H₂ Storage Detection.* AI porosity and storage quality detection in high-dimensional attribute space. Identifies porous, laterally continuous reservoir units for carbon capture and hydrogen storage. Scalable across onshore and offshore sites. (Reference site: Stavanger, Norway.) - *Geochemical Exploration.* Geology-respecting ML that converts legacy soil, regolith, and drillhole geochemistry into ranked, defensible targets. Spatially blocked validation. Robust on highly imbalanced exploration datasets. (Reference site: Pilbara, Western Australia.) **Operational intelligence — real-time decision support.** - *Well Monitoring Agent.* Continuous decision-support layer ingesting real-time drilling parameters, mud logs, wireline data, pressure readings, and integrity metrics. Detects anomalies and geological inconsistencies. Compares against historical failure modes. - *Exploration Signal Agent.* Monitors technical papers, government releases, reserve reports, and academic publications. Detects discovery indicators, reserve upgrades/downgrades, and changes in recovery factors. - *Discovery Intelligence Agent.* Synthesises discovery announcements from filings, regulatory disclosures, and industry publications. Classifies by materiality and strategic relevance. (Reference site: Ghawar, Saudi Arabia.) **Strategic intelligence — cross-sector agents.** - *M&A Pre-Signal Intelligence.* Detects early acquisition or divestment intent. Scans capital spending shifts, strategic reviews, advisor appointments, management language changes. - *ESG Risk & Opportunity.* Monitors environmental, social, and governance signals. Differentiates noise from structurally material issues. Assesses impact on permitting, timelines, capex, reputation. - *Capital Allocation & Capex.* Monitors shifts in corporate capital allocation. Translates corporate disclosures into forward-looking investment intelligence. - *Supply Chain & Critical Inputs.* Early-warning for production outages, export restrictions, permitting delays, logistics bottlenecks, force majeure. (Reference site: Port of Rotterdam.) - *Regulatory & Litigation.* Filters regulatory developments into decision-relevant intelligence. Monitors emerging legal disputes and litigation risks before they are priced in. **Delivery models.** - *Raw Outputs.* Seismic-compatible volumes, forecasts, and data files delivered directly to client interpretation systems. - *Models as a Service.* Hosted model access with API integration into existing client workflows. - *Consulting & Build.* Expert-led engagements for bespoke model design, analysis, interpretation, and strategic advisory. --- ## Industry — Planning **Positioning.** Purpose-built AI agent systems for planning consultancies, local authorities, and infrastructure delivery bodies. Automating repetitive tasks, enforcing compliance, and streamlining document workflows. **The challenge — an industry under pressure.** - *Talent shortage.* Acute planner shortage worldwide. University pipelines insufficient. Appeals and decisions taking 18+ months due to lack of experienced inspectors. - *Process bottlenecks.* A single missing annotation can fail validation and reset timelines by months. Wind farm radius checks take four full days. Public consultations generate hundreds of objections requiring individual review. - *Regulatory complexity.* EU directives transposed into national law, interpreted differently by each council. Planners cross-reference 10+ policy sources per application. Three extra sentences from a new directive can cause disruption. - *Security gaps.* Staff using personal ChatGPT accounts for professional work. Sensitive client data processed through unsecured channels with no audit trail, no GDPR compliance, no data residency controls. **Capabilities.** - *Document processing & generation.* Automated report generation enforcing house style and regulatory requirements. Validation agents catch errors before submission. Template management with regulatory layering that auto-propagates directive changes. - *Regulatory compliance monitoring.* Watch agents monitoring EU directives, national legislation, and council plans. Dynamic compliance checklists by project type and location. Policy cross-referencing across 10+ sources. Interpretive audit trails. - *Site feasibility & due diligence.* Automated screening aggregating flood risk, heritage, zoning, and servicing data. Infrastructure availability assessment. Visual constraint mapping cross-referenced against policy. (Reference site: Dublin, Ireland.) - *Application tracking & workflow.* Project lifecycle dashboards with real-time status. Milestone and deadline tracking with automated alerts. Stakeholder coordination. Validation pre-flight checks before submission. - *Data processing automation.* Cumulative impact assessments reduced from four days to minutes. Public submission processing — AI reads, categorises, and summarises hundreds of objection letters. Data extraction from council registers. - *Enterprise-grade security.* PII tokenisation proxy ensuring raw data never reaches AI providers. Per-tenant AES-256-GCM encryption. Three-layer PII detection. Full audit trails. Role-based access controls. **Engagement model.** 1. *Phase 1 — Discovery.* Embed with the planning team. Map workflows, pain points, templates, regulatory requirements. 2–4 weeks. 2. *Phase 2 — Pilot.* Deploy agents against highest-impact targets first: document generation, spreadsheet processing, compliance checklists. Full human oversight. 3. *Phase 3 — Scale.* Expand coverage across the workflow. Add feasibility automation, application tracking, submission processing. Platform integrations completed. (Reference site: Singapore.) 4. *Phase 4 — Optimise.* Ongoing monitoring, model updates, workflow refinement. **Market context.** Governments worldwide are committing record capital to housing, transport, energy, and water infrastructure. The planning workforce cannot scale to match — severe shortages of qualified planners reported across the UK, Ireland, EU, and beyond. The renewable energy pipeline (solar, wind, battery storage) is creating unprecedented application volumes with standardised regulatory requirements — a natural fit for AI-augmented workflows. --- ## Industry — Finance **Positioning.** Purpose-built AI agents for the private equity and financial services sector. Automate fund screening, portfolio management, fundraising, compliance, and deal execution. **LP-focused agents.** - *Mandate Screener.* Automate initial fund due diligence. Reads fund materials and ranks by mandate fit. 80% reduction in screening time. - *Pipeline Probability Scorer.* Predicts fund approval outcomes using historical decision data. 85%+ AUC-ROC accuracy. - *Vintage Comparator.* Objective fund benchmarking with IRR, TVPI, DPI, and PME. ILPA and GIPS compliant. - *Allocation Planner.* Quantitative pacing and liquidity management. Monte Carlo simulation for commitment sizing. - *Manager Risk Monitor.* 24/7 real-time surveillance on GP relationships. 48–72 hour early warning advantage. **GP-focused agents.** - *LP Targeting Engine.* Match fund strategy to 10,000+ institutional investor mandates. 30% improvement in meeting conversion. - *DDQ/RFP Writer.* Auto-draft questionnaire responses from a version-controlled knowledge base. >90% answer accuracy. - *Fundraising Momentum Tracker.* Real-time LP engagement scoring with reinforcement learning recommendations. - *Value Creation Advisor.* AI-powered portfolio company diagnostics. Root cause identification in <2 hours vs 2 weeks. - *Deal Screening Filter.* Triage inbound deals automatically. 10x pipeline throughput. **AML & compliance.** Automated sanctions screening, identity verification, and PEP detection across every counterparty and transaction. Suspicious activity flagged in real-time with explainable reasoning and recommended next steps. Every decision carries a full audit trail. GDPR-compliant from the ground up, configurable retention, right-to-erasure baked in. **Why Ziro (vs. traditional approach).** - Deployment: weeks vs. 6–18 months. - Architecture: modular, agent-by-agent vs. monolithic. - Explainability: built-in (SHAP, cited sources) vs. black box. - Compliance: GDPR-native, full audit trail. - Integration: CRM, data providers, accounting. - Learning: continuously improves vs. static. --- ## Platforms — overview SaaS products built for a specific industry. Off-the-shelf with institutional-grade depth — purpose-built for the domain, ready to deploy, no custom development required. - **Platform — Finance** (`/platforms/finance`): Quantitative trading and portfolio intelligence. - **Platform — Legal** (`/platforms/legal`): Contract lifecycle and case management. - **Document Editor** (https://docs.ziro.pub): AI-native document creation and collaboration. --- ## Platform — Legal **Positioning.** AI-native practice management. From inquiry to invoice, every step accelerated by intelligent agents that draft, classify, extract, and automate — while caseworkers retain full control. **Agents draft, caseworkers decide.** **Practice management capabilities.** - *AI-powered inquiry triage.* Incoming inquiries automatically classified by practice area with confidence scores. AI generates follow-up questions and drafts first-contact responses in the caseworker's tone. - *Document drafting engine.* Client Care Letters, witness statements, bills, file summaries, and correspondence drafted by AI agents. Every output preserves original DOCX template formatting. - *Tone-matched communication.* Each caseworker's writing style is profiled from sample emails. AI drafts match their formality, empathy, directness, and key phrases. - *Case lifecycle management.* Full status workflow from inquiry to closure. Assign primary and secondary caseworkers, track progress through configurable workflow checkpoints. - *Deadline extraction & alerts.* NLP extracts court dates, filing deadlines, tribunal listings, and limitation dates from emails. Warnings fire with configurable lead times. - *Chase-up automation.* Seven rule types evaluate case status on a schedule. CCL not returned? Documents missing? Filing deadline approaching? The system creates, tracks, and auto-resolves chase-ups. - *Time & billing pipeline.* Log time entries by type, auto-track client-facing activities, generate itemised bills with VAT. Draft → approval → sent → paid. - *Microsoft 365 integration.* Outlook inbox monitoring, email sending as the caseworker's account, and full OneDrive file management. Case folders created automatically. - *Audit & compliance.* Append-only audit logs with immutable history. SRA 6-year retention, GDPR-compliant UK data residency, before/after state capture for every mutation. **Workflow — from inquiry to invoice.** 1. *Inquiry arrives.* Email, phone, HubSpot, website form, or walk-in. Ingested and normalised regardless of source. 2. *AI classifies & drafts.* Practice area identified with confidence scores. Follow-up questions generated. First-contact response drafted in the caseworker's tone. 3. *Caseworker reviews.* Draft appears in queue. They edit in a clean text editor, approve, and send via Outlook as their own email. 4. *File opening.* One click triggers the full file-opening workflow. Client Care Letter, Letter of Authority, File Summary, and Case Data Sheet generated and queued for review. 5. *Case progresses.* Checkpoints define milestones. Deadlines extracted from correspondence. Chase-ups fire when overdue. Time entries accumulate. 6. *Billing & closure.* Itemised bill generated from time entries. VAT calculated. Sent and tracked to payment. Case closed with full audit trail. **Document engine — any document, fully bespoke.** Import existing DOCX templates or create new ones. AI drafts from your templates while preserving every formatting detail. Lawyers edit in a clean editor with full change tracking. Always your document, your formatting, your voice. - *Template system.* Upload firm's existing DOCX templates with placeholders. Platform populates with case data, client details, and AI-generated content. - *AI-assisted drafting.* Drafts in each caseworker's tone, generated from templates not from scratch. - *Import & edit any document.* Import any DOCX. Edit directly with clean text editor and full change tracking. Export to DOCX or pixel-accurate PDF. - *Bespoke document creation.* Create new templates directly or upload a Word file as a starting point. Configure placeholders, define which AI agents populate which sections. - *DOCX-to-PDF rendering.* Headless conversion pipeline. What the caseworker sees in preview is exactly what the client receives. - *Version control & audit.* Every edit, generation, and approval tracked. Full version history with before/after diffs. Ready for SRA inspection. **Access control — role-based permissions.** - *Caseworker.* Own drafts, cases, time entries, deadlines. Full editing and send permissions on assigned work. - *Senior Caseworker.* Everything a caseworker sees plus read-only visibility across the team. Monitor workloads and case progress. - *Admin.* Full visibility. Manage users, roles, templates, workflow configuration, system settings. **Intelligence — two pipelines, one intelligence layer.** The Compliance Pipeline checks the "what." The Legal Pipeline checks the "how" and "why." Together, they close the loop. - *Compliance Pipeline.* Models regulatory frameworks, controls, and procedures as a knowledge graph. Automatically identifies applicable frameworks, assesses controls, produces compliance scores (0–100), monitors regulatory sources every 6 hours, generates remediation plans. - *Legal Pipeline.* Models legislation, case law, legal principles, and procedures as a connected knowledge graph. Identifies applicable legislation, provides step-by-step procedural guides with fees and timelines, tracks legal interpretations through citation networks, monitors for amendments and new case law. **Compliance scoring tiers.** - *70–100: Proceed.* Controls in place and mapped. Project proceeds. - *50–69: Oversight.* Gaps identified. Remediation tasks created. Project proceeds with oversight. - *0–49: Remediate.* Critical gaps. Remediation required before proceeding. **Three value pillars.** - *Risk reduction.* Automated monitoring catches regulatory and legal changes before they become audit findings. Cross-framework gap analysis eliminates blind spots. Citation networks trace evolving interpretations. - *Operational efficiency.* Matter analysis in under two seconds. Step-by-step procedural guides with fees and timelines. Cross-framework control mapping reduces audit burden. - *End-to-end governance.* Full audit trail of every compliance check, legal assessment, review decision, and approval. Human-in-the-loop review on every output. Unified dashboard. Quantified scores for partners and regulators. **Practice areas (configured to the firm).** Immigration, Regulatory Compliance, Family, Employment, Corporate, Property. Additional practice areas configured based on the firm's operations. **What you get.** - *Full data intelligence.* Access to the compliance, legal, and regulatory database with the firm's own documents connected. Internal policies, precedent documents, and operational procedures ingested, embedded, and cross-referenced against both knowledge graphs. Continuous alerting when regulations change. - *Personalised platform.* Dedicated platform built on existing frameworks — firm's branding, domain, workflows. Full AI integration with intelligent interface. Intelligence embedded directly into daily tools. - *Integrated AI workflows.* End-to-end automated processes with human-in-the-loop review at critical decision points. Configurable workflows built around the firm's documents, obligations, and team structure. **Integration.** API-first architecture; agent connectivity; webhook alerts via email, Slack, Microsoft Teams; embedded intelligence in CRM, document editor, or project management tools; document ingestion; custom dashboards. **Next steps.** 1. *Discovery call.* Walk through the firm's current compliance and legal landscape, regulatory obligations, and operational workflows. 2. *Technical assessment.* Map practice areas, jurisdictions, client base, and operations to the platform's coverage. 3. *Proof of concept.* Run the firm's profile through both pipelines and review the automated assessment. 4. *Engagement proposal.* Recommended configuration and detailed implementation timeline. **Disclaimer.** This platform is a decision-support tool designed to assist legal and compliance professionals. It does not constitute legal advice and does not replace the judgement of qualified legal counsel. All AI-generated outputs pass through human review before action is taken. Scores are indicative assessments and should be validated by professional advisors. --- ## Platform — Finance **Positioning.** Trade with quant firm power. The first open-architecture quantitative trading ecosystem. Institutional infrastructure, no walled gardens, no black boxes. **Core capabilities.** - *Visual Strategy Builder.* Build sophisticated trading algorithms without writing boilerplate. Drag-and-drop for speed, or inject custom Python/JavaScript nodes for infinite flexibility. Live debugging. - *Machine Learning & Reinforcement Learning.* Deploy frameworks used by top quant firms — TensorFlow, PyTorch, Stable-Baselines3. RL agents to discover novel strategies. Automated hyperparameter tuning via Optuna. - *Institutional-Grade Backtesting.* Event-driven engine simulates decades of market data in minutes, accounting for slippage, transaction costs, and order book depth. Test 10,000 strategy variations in a single weekend. - *Open Microservices Architecture.* 13 independent, containerised microservices. Run only what you need. Deploy via Docker on local machine, AWS, or private VPS. Full REST and WebSocket APIs. - *Satellite-Guided Inference.* Harness alternative data streams from orbital imagery. Track retail foot traffic, shipping activity, agricultural yields in real-time. Computer vision pipelines transform raw satellite feeds into actionable trading signals. **The Ziro advantage (vs. traditional retail).** - Strategy creation: Full Python/JS + Visual Builder vs. limited scripting. - AI/ML: Full TensorFlow/PyTorch vs. non-existent or black box. - Data access: Open APIs vs. proprietary, locked-in. - Infrastructure: Private Docker, low latency vs. shared, high latency. - Testing cost: Zero (backtesting) vs. high (live required). **Built for you.** - *For the data scientist.* Institutional-grade infrastructure for your analysis. Explainable deep learning models with full auditability, no more black boxes. Focus on alpha generation, Ziro handles engineering complexity. - *For the developer.* Build trading systems the way you build software. Clean abstractions, version control, CI/CD, observability built in. Your engineering discipline finally applies to financial markets. - *For the discretionary trader.* Codify decades of pattern recognition into systematic strategies. Backtest against historical regimes, stress-test against black swans, execute with mathematical precision while preserving intuition. --- ## Contact & onboarding **Get in touch.** - Email: info@ziro.pub - LinkedIn: https://www.linkedin.com/company/ziroai/ - Location: 4th floor, 34–36 Grays Inn Rd, London, WC1X 8HR, United Kingdom **Onboarding process.** 1. *Submit your request.* Complete the comprehensive quote form with details about your organization and project needs. 2. *Consultation call.* Ziro reaches out to discuss requirements and understand your vision. 3. *Tailored proposal.* Receive a customised proposal outlining the recommended approach, timeline, and investment. **Quote form.** https://forms.ziro.pub — multi-step, save-and-continue via magic link. **Existing requests.** https://forms.ziro.pub/login — log in to view submissions, continue drafts, check status. --- ## Site map ``` / Home — positioning, philosophy, team, global presence /industries Industries listing — problems-vs-solutions explorer /industries/legal Agentic legal intelligence /industries/petroleum Subsurface intelligence /industries/planning AI-powered planning intelligence /industries/finance AI agents for private equity /platforms Platforms listing /platforms/legal AI-native practice management /platforms/finance Quant trading ecosystem /contact Contact form + onboarding process /form Multi-step quote form (also at forms.ziro.pub) /privacy Privacy policy /terms Terms of service ``` External platforms hosted under separate domains: https://docs.ziro.pub (Document Editor).