CUSTOMER SUCCESS STORY · FINANCIAL SERVICES
HD Securities expands GenAI from financial knowledge to internal AI assistants on AWS.
Generative AI · Financial RAG · Enterprise AIAT A GLANCE
The project at a glance.
HD Securities develops securities services and digital investing experiences while applying AI to financial knowledge and employee productivity. TitanBases supported two distinct engagements: a financial GenAI and RAG platform, and AiOffice (HDWork), an internal AI platform developed by FintechAI for HDS.
According to TitanBases, the SOW-defined engagements have been successfully delivered and are in production. The architecture combines protected AWS application and data foundations with Amazon Bedrock for generative AI and Bedrock Knowledge Bases for retrieval-augmented generation.
Solution implemented
- Financial GenAI and RAG with controlled source documents, Amazon S3, Titan Text Embeddings V2 and OpenSearch Serverless.
- Amazon Bedrock processing for AiOffice and HDWork, connecting AI agents to internal knowledge and existing AWS applications.
- Networking, data, access, encryption and observability controls appropriate for financial workloads.
Delivery outcomes
- Two distinct AI workstreams were successfully delivered, as confirmed by the project owner.
- Context-aware financial knowledge retrieval and managed model integration on AWS.
- A foundation for customer-facing AI experiences and internal employee productivity with operations and security controls.
ABOUT HD SECURITIES (HDS)
Digital finance powered by trusted knowledge and AI - for customers and employees.
HD Securities provides brokerage, proprietary trading, investment and financial advisory, custody and underwriting services. The company is developing digital investing and wealth-management experiences supported by technology, data and contextual customer service.
In parallel with investor-facing experiences, HDS uses AI for internal knowledge, drafting, summarization, analysis and workflows. AiOffice (HDWork), developed by FintechAI for HDS, was described in the v1.2 SOW as having more than 50 AI agents in development - not 50 verified production agents.
Digital securities
Brokerage, investment advice and wealth-management experiences.
Financial knowledge
Grounded retrieval from controlled financial knowledge sources.
AI customer assistance
Help with products, procedures and investment-service questions.
AiOffice · HDWork
AI-assisted documents, analysis and internal workflows.
TitanBases collaborated on two distinct technical scopes: HDS financial GenAI and RAG, and AiOffice or HDWork developed by FintechAI. They have separate architectures and responsibilities, so this case study presents them independently rather than merging them into one system.
THE CHALLENGE
Business challenge · Financial knowledge · Enterprise productivity
Scaling AI in financial services takes more than foundation-model access.
HDS needed grounded answers to questions about products, accounts, trading and financial topics using controlled sources. Employees also needed AI-assisted drafting, summarization, knowledge access, data analysis and workflow automation.
Financial RAG required protected knowledge retrieval and a reliable application layer. AiOffice already had application infrastructure on AWS but used Google Vertex AI for model services; the migration workstream required evaluation of Amazon Bedrock integration, model suitability, reliability and governance.
TitanBases owned infrastructure, model integration, knowledge retrieval, security and validation within the SOWs. HDS and FintechAI retained responsibility for application code and business logic.
ENGINEERING AT SCALE · TWO AI WORKSTREAMS
One financial institution. Two AI journeys. Controlled AWS foundations.
Protected application and data layers connect to retrieval-augmented generation and internal AI workflows, while each engagement retains its own SOW-defined system boundaries.
Delivery and production status are confirmed by TitanBases. Design figures are not treated as measured operating KPIs or production agent counts.
THE SOLUTION
AWS architecture · Financial RAG · AiOffice
Two specialized AI workloads, implemented on AWS.
For financial GenAI and RAG, application traffic passes through Route 53, AWS WAF and load-balanced private compute. Source knowledge is ingested from Amazon S3, embedded with Titan Text Embeddings V2 and indexed in OpenSearch Serverless. Bedrock Knowledge Bases retrieves context for model-generated responses; PostgreSQL supports application data.
For AiOffice (HDWork), FintechAI develops the internal AI application on AWS. TitanBases supported the Bedrock model-processing and migration workstream. The SOWs differ in their compute descriptions; both source diagrams remain separate and must be checked against the latest as-built state before public disclosure.
SOLUTION ARCHITECTURE
AWS Architecture · Two distinct workstreams
ORIGINAL DIAGRAMS FROM HDBS v1.0 AND AIOFFICE v1.2 SOWS
01 · Financial GenAI / RAG
02 · AiOffice / HDWork
Route 53, WAF, load balancing and private networking.
Application workloads and PostgreSQL data services across controlled AWS environments.
Amazon S3, embeddings, OpenSearch and Bedrock Knowledge Bases.
Amazon Bedrock integration, Guardrails and bounded error handling.
DELIVERY SUCCESS · PRODUCTION
Two generative AI engagements, delivered and brought into operation.
TitanBases confirms completion and production status. The following are delivered technical capabilities and workloads - not invented model-accuracy, employee-time-savings or measured-uptime statistics.
FINANCIAL KNOWLEDGE
RAG
Grounded financial-knowledge retrieval that augments context before response generationGENAI MODEL PLATFORM
Bedrock
Managed foundation-model access integrated with the existing application environmentENTERPRISE PRODUCTIVITY
HDWork
Document work, internal knowledge and employee-oriented AI workflowsSECURITY · OBSERVABILITY
Secure
Access, encryption, observability and repeatable infrastructure practicesThe 99.9% availability and concurrency criteria in the SOWs are design and test targets, not published production measurements. More than 50 AiOffice agents are described as in development, not as verified live agents.
TECHNOLOGIES
An AWS ecosystem for secure applications, data and AI.
Services are combined from two separate SOWs and grouped by their roles. This does not imply that every listed service runs in the same HDS production cluster or environment.
APPLICATION AND NETWORKING
GENERATIVE AI AND RAG
DATABASES AND OPERATIONS
ENGINEERING AND GOVERNANCE
PRODUCTION · NEXT STEPS
From financial retrieval to AI that supports everyday work.
TitanBases confirms both projects were delivered. With HDS approval and current as-built and operational evidence, this story can incorporate response times, RAG quality, productivity improvements and AiOffice rollout scope without changing the Golden Visual.
CUSTOMER STORY · FINANCIAL SERVICESRELATED CAPABILITIES
RELATED CASE STUDIES
Explore more stories.
RELATED INSIGHTS
engineering insight
AI for Financial Services in Vietnam: TitanBases Visits FintechAI and AWS in Ho Chi Minh City
A working visit to FintechAI in Ho Chi Minh City gave TitanBases a closer view of how AI initiatives in Vietnam's financial-services sector are evolving from individual use cases toward production-ready systems, data platforms, infrastructure, and enterprise integration.
Sep 25, 2026
engineering insight
MongoDB Atlas Vector Search for Large-Scale eKYC Deduplication: Lessons from a 10M-Vector POC
Large-scale identity matching requires more than fast vector search. This engineering insight examines latency, throughput, recall, index behavior and quantization trade-offs observed while evaluating MongoDB Atlas Vector Search for a 10-million-vector eKYC workload.
Sep 21, 2026
engineering insight
Buying ChatGPT Business in Vietnam: A Procurement Checklist for Teams
A practical procurement guide for teams in Vietnam comparing ChatGPT Business Standard and Premium pricing, seat planning, signup steps, privacy controls and onboarding support.
Nov 5, 2026
Prepared by TitanBases Engineering Team
TitanBases engineers share practical guidance from cloud, data and AI delivery work, with technical review focused on production architecture, security and operations.
YOUR NEXT WORKLOAD