Senior Technical Writer · Documentation Architect · FinTech Specialist
20 years architecting enterprise documentation systems for high-stakes domains: payments, fraud prevention, financial compliance, and investment management. I build documentation infrastructure that scales with complexity—turning complicated systems into clear, maintainable knowledge.
Expertise: API References (Stripe/Plaid quality) · FinTech Domain Depth · Docs-as-Code Automation · Compliance Frameworks · System Architecture
Current Focus: Open to senior-level opportunities in tech companies looking to scale documentation systems and establish documentation excellence.
🌐 Portfolio & profile: sulagnasasmal.com
- 📝 Developer API references — REST endpoints, webhooks, OAuth, error codes (Stripe / Plaid quality bar)
- 🏛️ Enterprise admin docs — MadCap Flare HTML, regulated platforms, step-by-step procedures
- ⚙️ Docs-as-Code pipelines — Vale linting, GitHub Actions, MkDocs, GitHub Pages
- 🤖 AI-assisted documentation tools — Claude API, Next.js + GPT-4o, RAG chatbots, docs-as-code frameworks
- 🔌 Developer portal tooling — OpenAPI spec rendering, interactive API consoles
- 💳 FinTech domain depth — ACH, Fedwire, SWIFT, RTP/FedNow, OFAC, BSA/AML, ISO 20022
| Project | Type | |
|---|---|---|
| FraudShield AI Engine Documentation | AI fraud detection · risk models · API · explainability | View → |
| PayPlus REST API Developer Reference | Developer API reference | View → |
| PayPlus Enterprise Administration Guide | Enterprise admin docs | View → |
| US Payments Hub | Operations & compliance reference | View → |
| VaultPay API Docs | Fintech payment API reference | View → |
| CaseForge API Docs | Enterprise AML & SAR filing API | View → |
| CaseForge Onboarding Kit | Post-sale customer onboarding · 30-day program · role-based checklists · go-live validation | View → |
| CaseForge SDK Documentation | Python · Node.js SDK · authentication · error handling · migration guide · versioned changelog | View → |
| Documentation Center Platform | Portfolio hub · writing philosophy · information architecture · doc-as-code workflow | View → |
| DocForge Platform | Docs-as-Code platform dashboard | View → |
| Docs-as-Code Portal | DaC methodology documentation | View → |
| RepoDocs AI | Claude API · docs-as-code framework for SaaS API teams · reusable templates · AI prompt library · GitHub Actions validation · OpenAPI-to-Markdown generation | View → |
| DocCraft AI | Next.js + GPT-4o docs generator | View → |
| SpecFlow | OpenAPI spec → interactive developer portal | View → |
| DocQuery | RAG-powered documentation chatbot | View → |
| DocPulse | Documentation analytics dashboard | View → |
| PPT → MP4 Automation | Python narrated video pipeline | GitHub → |
| AI & Automation Tools Portal | All AI tools portal — RepoDocs AI, DocCraft, SpecFlow, DocQuery, DocPulse, PPT→MP4 | View → |
| SunBridge Asset Atrium Manager — Platform Docs | Investment platform architecture · NAV batch · HA · multi-TZ · release governance | View → |
| Asset Atrium Compliance Framework | Investment compliance · pre-trade · post-trade · mandates · UCITS · AIFMD · MiFID II | View → |
| NexaFlow SDK Documentation | SDK & code library docs · API reference · quickstart · error handling · migration guide · changelog | View → |
| Docs Content Strategy & Style Guide | IA framework · voice & tone · doc types · audience analysis · metrics · templates · audit tools | View → |
| US ACH Payment Processing Documentation | End-to-end ACH payment lifecycle · origination · returns · NOCs · compliance · developer integration | View → |
| DOC Portal | Curated portfolio of API references, admin guides, knowledge bases & AI-powered documentation tools | View → |
| Developer Onboarding — Payments API | Zero-to-live-payment in 5 steps — account, SDK, test calls, webhooks, go-live checklist | View → |
| FinTech Learning Academy | Structured courses on payments, fraud detection, ACH processing & compliance | View → |
| AML Compliance Training Program | Annual BSA/KYC/CDD/OFAC screening & transaction-monitoring training for FinTech | View → |
| Payments Integration Tutorials | Hands-on developer tutorials — ACH transfers, webhooks, sandbox testing & API auth | View → |
|
|
Post-sale onboarding package for enterprise CaseForge AML customers. 30-day time-boxed program with role-based checklists, environment provisioning guides, data migration procedures, user training tracks, and a go-live validation checklist.
Language: HTML | Live Site: sulagnasasmal.github.io/caseforge-onboarding | Repo: github.com/SulagnaSasmal/caseforge-onboarding
Onboarding · Enterprise AML · Checklists · Go-live Validation · Role-based Tracks
Python and Node.js SDK guide for the CaseForge AML REST API. Covers quickstart, authentication (API keys + OAuth 2.0), a structured exception hierarchy, full code-sample library, versioned changelog, and a v1.x → v2.0 migration guide with side-by-side breaking-change comparisons.
Language: HTML | Live Site: sulagnasasmal.github.io/caseforge-sdk-docs | Repo: github.com/SulagnaSasmal/caseforge-sdk-docs
SDK Documentation · Python · Node.js · OAuth 2.0 · Migration Guide · Versioned Changelog
Capstone portfolio hub bringing all 7 documentation phases together. Includes meta-documentation on writing philosophy (MSTP compliance, before/after examples), information architecture decisions per phase, and a full doc-as-code workflow with branch strategy, Vale linting, GitHub Actions CI, and deployment pipeline.
Language: HTML | Live Site: sulagnasasmal.github.io/Documentation-Center-Platform | Repo: github.com/SulagnaSasmal/Documentation-Center-Platform
Portfolio Hub · MSTP · Information Architecture · Doc-as-Code · Vale Linting · GitHub Actions
Full developer reference for a workflow automation SDK — installation, core concepts, authentication, error handling with retry policies, full API reference (client, workflows, triggers, actions), v1-to-v2 migration guide, and versioned changelog. Written to the same quality bar as Stripe and Twilio SDK docs.
Language: Markdown | Repo: github.com/SulagnaSasmal/nexaflow-sdk-docs
SDK Documentation · API Reference · Migration Guide · Error Handling · Node.js · Python
A complete documentation program framework: editorial style guide (voice & tone, formatting, naming, code examples), information architecture methodology, doc type taxonomy, audience analysis framework, metrics program, four ready-to-use templates, a doc audit checklist, and a content health scorecard. Built for teams managing documentation at scale.
Language: Markdown | Repo: github.com/SulagnaSasmal/docs-content-strategy
Content Strategy · Style Guide · Information Architecture · Docs Metrics · Templates · Content Audit
HTML: ██████████████░░░░░░ 69% TypeScript: ██░░░░░░░░░░░░░░░░░░ 9% JavaScript: █░░░░░░░░░░░░░░░░░░░ 6% Python: █░░░░░░░░░░░░░░░░░░░ 6% CSS: █░░░░░░░░░░░░░░░░░░░ 6% MDX: █░░░░░░░░░░░░░░░░░░░ 3%
| Repository | Last Update | Recent Commits |
|---|---|---|
SulagnaSasmal |
4/13/2026 | 3 commits |
my-docs-practice |
4/9/2026 | 3 commits |
blog |
4/1/2026 | 3 commits |
repodocs-ai |
4/1/2026 | 3 commits |
sulagnasasmal-site |
3/31/2026 | 3 commits |
ai-doc-tools |
3/30/2026 | 3 commits |
doccraft-help-center |
3/28/2026 | 3 commits |
documentation-ai-agent |
3/27/2026 | 3 commits |
doc-portal |
3/20/2026 | 3 commits |
docs-content-strategy |
3/19/2026 | 3 commits |
A visitor analytics system deployed on Vercel tracks page interactions across this GitHub profile in real time.
This profile includes a comprehensive visitor analytics system that tracks how people interact with this GitHub presence. It provides insights into:
- 👁️ Page Views — Profile visit count, referrer sources, device types, geographic data
- 🖱️ Click Tracking — Every link and button clicked, with coordinates and scroll depth
- 📜 Behavior Patterns — Time on page, scroll depth, session duration, interaction count
- 📱 Device Distribution — Desktop vs. mobile vs. tablet breakdown
- 🔗 Top Content — Most-clicked links and most-engaged projects
The analytics infrastructure includes:
| Component | Purpose |
|---|---|
| Tracking Script | Client-side JavaScript that collects visitor data in real-time |
| Analytics Server | Node.js backend storing events in JSONL format by date |
| Dashboard Generator | Auto-generates HTML dashboards and Markdown reports |
| Metrics Collection | Weekly aggregation of GitHub stats, followers, stars |
| Historical Tracking | 90-day rolling analytics history with growth trends |
Visit Profile
└─► Tracking script loads
└─► Records: page view, device, referrer, timestamp
└─► Events sent to analytics backend
└─► Stored in analytics-data/{projectId}/{date}/*.jsonl
└─► Dashboard generated from aggregated data
└─► Markdown report & HTML dashboard available
The system collects only non-personal analytics data:
- Session ID (anonymous token, not connected to identity)
- Device type & screen resolution
- Browser & OS information
- Referrer source
- Links clicked and interaction coordinates
- Scroll depth and time on page
- Page visit timestamps
✅ No personal data collected (no names, emails, or IP addresses logged) ✅ Fully anonymized sessions — Session IDs are random, not linked to identity ✅ Optional tracking — Can be disabled via configuration ✅ Local storage first — Events stored locally before sending ✅ User data control — Data can be cleared from browser localStorage
For local testing:
# Start the analytics server
node scripts/analytics-server.js
# Server runs on http://localhost:3001
# Generate dashboard
node scripts/dashboard-generator.js
# Output: analytics-dashboards/index.html & ANALYTICS_REPORT.md
# Collect metrics
npm run collect-metrics
# Output: metrics/analytics.json with 90-day history📖 Full guide: See docs/ANALYTICS_GUIDE.md for implementation details, deployment options, and troubleshooting.
Each project below is a fully authored, end-to-end documentation site — built using Docs-as-Code principles or enterprise HTML, hosted on GitHub Pages.
A light-theme personal portfolio showcasing my work, background, and expertise.
A hand-coded personal portfolio site — single HTML file, no frameworks. Covers my professional background, featured documentation projects, compliance expertise, experience timeline, and a personal section on travel and food.
Light theme · Amber accent (#E8A84C) · Inter + Outfit + JetBrains Mono · No frameworks · GitHub Pages
Enterprise-grade REST API reference for a fintech payment processing platform, aligned to NICE Actimize and global financial compliance standards.
A comprehensive REST API reference for VaultPay — a realistically modelled fintech payment platform. Documents the complete payments lifecycle: authentication, settlement, disputes, and compliance reporting. Structured to match the quality bar of real-world developer portals (Stripe, Adyen, Braintree).
| Audience | What they find here |
|---|---|
| Backend developers | Complete endpoint reference, request/response schemas, curl examples, SDKs |
| Compliance teams | PCI DSS, PSD2/SCA, BSA/AML, GDPR coverage, AML alert management |
| Security engineers | HMAC-SHA256 webhook verification, 3D Secure flow, SCA exemption handling |
PCI DSS Level 1 · PSD2 / SCA · BSA/AML · FATF · NICE Actimize: IFM-AI · SAM · CDD · Sanctions
API reference for an enterprise AML/compliance case management platform — SAR filing, tamper-evident audit trails, regulatory workflow automation.
A v10.2 API reference for CaseForge — an enterprise compliance platform for managing AML investigations, filing SARs with FinCEN, and maintaining tamper-evident audit trails. Written for compliance officers, legal teams, and system integrators.
| Audience | What they find here |
|---|---|
| Compliance officers | SAR lifecycle, FinCEN BSA e-filing integration, case status workflows |
| System integrators | Full API reference, authentication, sandbox credentials, webhook payloads |
| AML analysts | Alert ingestion, case assignment, escalation paths, goAML export |
| Legal / Audit teams | Tamper-evident audit trail, SHA-256 hash chaining, signed PDF export |
FinCEN BSA e-filing · FATF Recommendation 20 · EU AMLD6 · SOC 2 Type II · ISO 27001
An interactive demonstration of a complete Docs-as-Code platform — from Markdown authoring through CI/CD validation to multi-format publishing.
A fully interactive Docs-as-Code platform dashboard — a single-page application with six modules demonstrating the complete documentation engineering pipeline.
| Module | What it does |
|---|---|
| Dashboard | Build health, content coverage, contributor metrics, pipeline run history |
| Live Editor | Split-pane Markdown editor with real-time preview and inline Vale linting |
| Content Map | Visual doc tree — spot orphaned pages, broken refs, coverage gaps |
| Style Linter | Vale rule enforcement: passive voice, banned phrases, readability, terminology drift |
| Build Outputs | View and download HTML site, PDF manual, OpenAPI spec, search index per build |
| Configuration | Site metadata, nav structure, theme tokens, lint rules, CI triggers, deploy targets |
Docs-as-Code · CI/CD for documentation · Vale style linting · Multi-format output · GitHub Pages
A full documentation portal explaining the Docs-as-Code methodology — what it is, how the pipeline works, and how to use the platform.
A 16-page documentation portal that explains the Docs-as-Code methodology, the five-stage pipeline, and how the Documentation Center platform works. Built like real product documentation — left nav, table of contents, code examples, step-by-step guides, callouts, comparison tables, and prev/next navigation. Self-referential: the portal is itself built using Docs-as-Code principles.
Documentation portal design · 16-page structured content · GitHub Actions CI/CD · Vale linting · MkDocs
An enterprise-grade reference site covering all six US payment rails — written for bank operations teams, compliance analysts, and payments professionals.
An eight-page HTML site covering ACH, Fedwire, SWIFT, RTP, FedNow, and Card Networks — the full landscape of US payment rails. Documents how each rail works, who owns it, message standards, key compliance requirements, and 12 real-world payment scenarios. Draws directly on my Fundtech India experience (2008–2010) documenting CashIn and Global CASHplus, the direct predecessors of Finastra GlobalPAYplus.
| Audience | What they find here |
|---|---|
| Bank operations staff | How each rail works, cutoff times, settlement windows, failure handling |
| Compliance analysts | OFAC, BSA Travel Rule, AML obligations by rail |
| Payments engineers | Message standards (ISO 20022, NACHA, SWIFT MT/MX), routing and BIC lookup |
| Payments onboarding teams | 12 scenario walkthroughs across all rails |
ACH · Fedwire (ISO 20022 March 2025) · SWIFT (MT + MX + gpi) · TCH RTP · FedNow · Card Networks · 12 scenario walkthroughs
An eight-section system administration guide for a fictional enterprise payment hub — modeled on the GlobalPAYplus class of multi-rail payment platforms.
A full system administration guide for PayPlus Enterprise v3.2 — a realistically modeled multi-rail payment hub for financial institutions. Covers installation, RBAC/LDAP/SSO, payment rail connector configuration, workflow engine, monitoring, OFAC/BSA/AML compliance controls, and troubleshooting.
| Audience | What they find here |
|---|---|
| System administrators | Installation, Java/DB requirements, post-install validation, upgrade procedure |
| IT security teams | RBAC model, LDAP/AD integration, SAML 2.0 SSO, password policy, session controls |
| Payments operations | Connector config per rail, approval workflows, STP rules, bulk ACH |
| Compliance officers | OFAC screening engine, hold queue actions, BSA Travel Rule by rail, AML integration |
Installation · User Management · Payment Rail Connectors · Workflow Engine
Monitoring & Alerts · Compliance Configuration · Troubleshooting
RBAC · LDAP / SAML 2.0 SSO · ISO 20022 pacs.008/pacs.009 · OFAC sanctions screening
BSA Travel Rule (31 CFR 103.33) · AML integration (NICE Actimize / Oracle FCCM) · MadCap Flare-inspired HTML
A six-section developer API reference for the same fictional payment hub — structured to the quality bar of leading fintech developer portals.
A developer-facing REST API reference for PayPlus REST API v2.1. Written to match the quality bar of leading fintech developer portals (Stripe, Plaid) — authentication, ACH payments, instant payments (RTP/FedNow), wire transfers (Fedwire/SWIFT), webhooks, and a complete error code reference.
| Audience | What they find here |
|---|---|
| Integration developers | OAuth 2.0, endpoint reference, request/response schemas, JSON examples |
| Core banking engineers | Idempotency, async ACH/wire pattern, synchronous instant payment response |
| Security engineers | HMAC-SHA256 webhook verification with replay protection |
| Compliance teams | OFAC error handling, BSA Travel Rule required fields, ACH return codes |
Authentication · ACH Payments · Instant Payments (RTP/FedNow) · Wire Transfers
Webhooks · Error Codes
OAuth 2.0 · Webhook HMAC-SHA256 · ISO 20022 (pacs.008, camt.056) · ACH return codes R01–R85 · SWIFT gpi / UETR
Enterprise documentation for an AI-powered fraud detection platform — risk scoring models, REST API reference, model explainability, and regulatory compliance.
A 9-section enterprise HTML documentation site for FraudShield AI Engine — a fictional AI fraud detection platform grounded in real NICE Actimize IFM/AMP architecture. Demonstrates expertise at the intersection of fraud domain knowledge, AI/ML model documentation, REST API reference writing, and financial regulatory compliance.
| Audience | What they find here |
|---|---|
| Fraud Operations analysts | Risk scoring model, score ranges, threshold tuning, false positive handling (suppression, whitelisting, step-up challenge flow) |
| Data Scientists / Model Owners | Model input features (200+ Risk Indicators), champion/challenger retraining cycle, drift triggers, training data requirements |
| Integration Engineers | Full REST API reference — POST /score, step-up result, case feedback — with JSON request/response examples, error codes, rate limits |
| Compliance Officers | Audit log schema, model explainability output, SR 11-7, GDPR Article 22, CFPB adverse action, BSA/AML SAR tagging |
Risk Scoring Model · Model Input Features · System Configuration · Threshold Tuning
False Positive Handling · Model Retraining Cycle · API Integration · Audit & Explainability
AI / ML Models · Risk Indicators (200+) · Champion/Challenger Framework · GDPR Art. 22
SR 11-7 Model Risk · BSA/AML · CFPB Adverse Action · REST API Reference · MadCap Flare-inspired HTML
Technical documentation for an enterprise investment management platform — J2EE architecture, NAV batch processing, high availability, multi-time zone fund operations, and release governance. Follows Microsoft Style Guide (MSTP) throughout.
Platform architecture and operations reference for SunBridge Asset Atrium Manager — a realistic J2EE institutional investment management platform modelled on the SunGard Asset Arena class of buy-side systems. Fully MSTP-compliant writing throughout: active voice, second person, sentence-case headings, no Latin abbreviations.
| Audience | What they find here |
|---|---|
| Platform engineers | WebLogic cluster design, Oracle RAC, IBM MQ Series, SWIFT Alliance, Bloomberg B-PIPE/BVAL |
| Operations / SRE | NAV batch orchestration, processing zones, HA design, rolling restarts, DR failover |
| Release managers | CAB governance, operational impact, backward compatibility, risk framing |
| Solution architects | J2EE component design, JTA transactions, Oracle Coherence cache, FIX 4.2/4.4 gateway |
Platform Overview · System Architecture · Batch Processing Framework
High Availability · Multi-Time Zone Processing · Failover & Audit Design
Database & Configuration · Release Impact Summary · Diagram Library
J2EE / WebLogic 14c · Oracle 19c RAC · IBM MQ Series · SWIFT Alliance Lite2 · Bloomberg B-PIPE
NAV batch processing · Oracle Coherence · FIX 4.2/4.4 · Mermaid diagrams · MSTP
Investment compliance documentation for an enterprise buy-side platform — pre-trade checking, post-trade monitoring, investment mandate management, exposure limits, regulatory reporting, breach governance, and audit trails. Follows Microsoft Style Guide (MSTP) throughout.
A structured compliance documentation set for Asset Atrium Manager — covering the full investment compliance lifecycle. Demonstrates buy-side domain expertise across investment mandate modeling, UCITS/AIFMD/MiFID II regulatory frameworks, breach management workflows, and compliance audit architecture. Fully MSTP-compliant writing throughout.
| Audience | What they find here |
|---|---|
| Portfolio managers | Pre-trade compliance (hard/soft breaches, override workflow), block order compliance |
| Compliance officers | Post-trade monitoring, breach lifecycle, escalation matrix, regulatory notification rules |
| Fund administrators | UCITS 5/10/40, AIFMD Annex IV, Form PF, MiFID II RTS 25, exposure & concentration limits |
| Operations / Audit | Oracle compliance audit schema, retention policy, Splunk integration, examination readiness |
Overview & Framework · Pre-Trade Compliance · Post-Trade Monitoring
Investment Mandates · Exposure & Concentration · Regulatory Reporting
Breach Management · Audit Trail & Logging
UCITS (5/10/40) · AIFMD Annex IV · MiFID II / RTS 25 · Form PF · Solvency II · EMIR · CPO-PQR
Pre-trade / Post-trade compliance · Breach lifecycle · Oracle audit schema · MSTP
Beyond documentation, I build automation tools that solve real workflow problems.
A full-stack AI application that converts raw content into polished, MSTP-compliant documentation — with auto-compliance checking, one-click fixes, and context-aware generation.
Feed DocCraft raw content — meeting notes, specs, screenshots, anything — and it produces publication-ready documentation. It thinks like a senior technical writer: analyzing gaps, asking clarifying questions, then generating structured output aligned to your existing style.
Upload raw content + context docs (style guide, glossary, OpenAPI spec)
└─► AI gap analysis → clarifying Q&A loop
└─► GPT-4o generates structured documentation
└─► MSTP compliance check runs automatically
└─► One-click Fix per issue (instant replace or AI rewrite)
└─► Export: HTML · Markdown · PDF
| Feature | What it does |
|---|---|
| Context Layer | Upload previous docs, glossaries, style guides, or OpenAPI specs — AI writes consistently with your existing content |
| MSTP Compliance Mode | Auto-runs after generation — checks forbidden words, passive voice, Title Case headings, non-imperative steps |
| One-click Fix | Instant string replace for terminology; AI-assisted rewrite for voice/structure issues |
| Inline AI editing | Select any text → Simplify, Expand, Add Example, Make Concise |
Next.js 14 · TypeScript · Tailwind CSS · OpenAI GPT-4o / GPT-4o-mini · React Markdown · Vercel
Central landing page for all AI-powered and automation tools — browse, explore, and link to DocCraft AI, SpecFlow, DocQuery, DocPulse, and the PPT→MP4 pipeline.
A single-page portal that surfaces all five AI and automation tools in one place — each with a live link, tech stack, and description. Clean navigation, consistent branding with the main portfolio site, and direct links to every tool's GitHub repo and live deployment.
HTML · GitHub Pages · Amber + Indigo + Teal theme · Responsive design · No frameworks
An end-to-end pipeline that converts PowerPoint presentations into narrated MP4 videos — no screen recording, no manual voiceover.
Fully automated pipeline: upload a .pptx → get back a narrated .mp4. No manual recording, no screen capture.
PowerPoint file
└─► Extract slide notes (python-pptx)
└─► Generate AI narration (Azure TTS — Jenny Neural)
└─► Export animated video (PowerPoint COM)
└─► Mux audio + video (FFmpeg)
└─► Download final.mp4
Python · FastAPI · Azure Text-to-Speech · PowerPoint COM · FFmpeg · pywin32
A Next.js application that transforms any OpenAPI 3.x specification into a fully interactive, Stripe-quality developer portal — with a live Try-It Console, authentication panel, and webhooks explorer.
Upload or paste any OpenAPI 3.x spec (YAML/JSON) and instantly get a fully rendered developer portal. Features grouped endpoint navigation, request/response schema trees with expand/collapse, an interactive Try-It Console for live API calls, authentication panel (API Key, Bearer, OAuth2, HTTP Basic), server selector, webhook payload explorer, and a search bar.
OpenAPI spec (YAML/JSON)
└─► Parse & validate (swagger-parser)
└─► Render grouped endpoints with schema trees
└─► Try-It Console with live API execution
└─► Auth panel + Server selector + Webhooks
Next.js 14 · TypeScript · Tailwind CSS · swagger-parser · Monaco Editor · Interactive API Console
A RAG-powered chatbot that lets users ask natural-language questions against documentation — with source citations, confidence scoring, and automatic content gap detection.
A complete RAG pipeline: crawl documentation sites, split content using section-aware chunking (respects heading hierarchy, never splits code blocks), embed with OpenAI text-embedding-3-small, store in ChromaDB, and answer questions with GPT-4o-mini citing exact sources. The Next.js chat frontend features message bubbles, source citation cards, confidence badges (green/amber/red), collection selector, and dark/light mode. Low-confidence answers are automatically logged as content gaps.
Doc sites → Crawl → Section-aware chunking → Embed (OpenAI)
└─► ChromaDB vector store
└─► Query → Retrieve → GPT-4o-mini → Answer + Citations
└─► Low confidence? → Auto-log content gap
| Component | What it does |
|---|---|
| Ingestion | Async web crawler + section-aware chunker + OpenAI embeddings → ChromaDB |
| Query Pipeline | Vector retrieval + confidence scoring (HIGH/MEDIUM/LOW) + conversation memory |
| Chat Frontend | Next.js UI with source cards, confidence badges, collection selector, dark mode |
| Gap Tracker | Auto-detects low-confidence answers and surfaces documentation gaps |
Python · FastAPI · LangChain · ChromaDB · OpenAI · Next.js 14 · TypeScript · Tailwind CSS
A lightweight analytics platform for documentation sites — tracks page views, reading time, search queries, content freshness, and feature coverage with a full Next.js dashboard.
A complete documentation analytics platform with two parts: a lightweight JavaScript tracker (<5KB, no cookies, anonymous sessions, batched events) that embeds in any doc site, and a FastAPI + SQLite backend with a Next.js dashboard. Dashboard includes an overview with metric cards and time-series charts, search analytics (top queries, failed searches), GitHub-based content freshness scoring, feature-to-documentation coverage matrix, feedback collection, and weekly report generation.
Doc site → tracker.js (pageview, scroll, search, feedback)
└─► Batch events → FastAPI → SQLite
└─► Next.js dashboard: Overview · Search · Freshness · Coverage · Reports
| Dashboard Page | What it shows |
|---|---|
| Overview | Page views, unique visitors, reading time, top pages, time-series charts |
| Search Analytics | Top queries, failed searches, content gaps from search |
| Freshness | GitHub-based staleness scoring (fresh/aging/stale/abandoned) |
| Coverage Matrix | Feature × endpoint × doc page mapping with status badges |
| Reports | One-click weekly performance report (Markdown) |
JavaScript · Python · FastAPI · SQLite · Next.js 14 · Recharts · Tailwind CSS
| Category | Tools & Technologies |
|---|---|
| Documentation formats | Markdown, MDX, OpenAPI 3.1, DITA concepts, reStructuredText |
| Docs platforms | MkDocs, Sphinx, Hugo, Docusaurus, ReadTheDocs, Confluence |
| CI/CD & automation | GitHub Actions, Vale linting, link checkers, build pipelines |
| API documentation | REST APIs, webhooks, request/response schemas, curl examples, SDK guides |
| Compliance domains | PCI DSS, BSA/AML, FinCEN, FATF, GDPR, SOC 2, ISO 27001, PSD2/SCA |
| NICE Actimize suite | IFM-AI (fraud), SAM (AML), CDD (KYC), Sanctions Screening |
| Dev tools | Git, GitHub, VS Code, Python, HTML/CSS, Bash, TypeScript, Next.js |
| AI & automation | OpenAI GPT-4o, Azure TTS, FastAPI, FFmpeg, GitHub Copilot, Claude, Napkin AI |
| Methodologies | Docs-as-Code, structured authoring, topic-based writing, information architecture, MSTP |
If you're looking for a Technical Writer who thinks like an engineer and writes for developers, compliance teams, and business stakeholders alike — let's talk.



