About
About Gokulakrishnan R
AI/ML & Generative AI Solutions Architect · Singapore. From Windows device drivers to production AI agents.
I'm a Technical Architect with 20+ years of experience architecting and delivering secure, scalable enterprise platforms, with a growing specialization in AI/ML and Generative AI. I build agentic AI and RAG applications using LangChain, LangGraph and Microsoft Foundry, develop ML models with TensorFlow, PyTorch and Scikit-learn, and design intelligent document processing pipelines that combine OCR, computer vision and knowledge graphs (Neo4j). My cloud background is deep on AWS, complemented by working knowledge of Azure AI/ML services (Azure ML, Azure OpenAI, Azure DevOps, Microsoft Fabric).
I own architecture decisions end-to-end, from model and framework selection through RAG/agentic design, deployment, security and performance, while working with cross-functional teams to deliver business-focused AI solutions in regulated, high-traffic environments. I also integrate AI capabilities into enterprise applications via REST APIs and automate multi-step workflows with tools like n8n.
Journey
From device drivers to AI agents
-
01
Systems & embedded
Windows device drivers, media-player firmware, a UPnP streaming server and a signature-less antivirus engine in C/C++.
-
02
Enterprise & retail
Voucher and store back-office systems for convenience, pharmacy and supermarket chains on Java and PostgreSQL.
-
03
Cloud SaaS & government
Multi-tenant SaaS for clients in the UK, Europe and Asia, a blockchain coupon platform, then serverless platforms for government and aviation.
-
04
Agentic AI
AI agents, RAG, MCP servers and computer vision, designed and shipped into production.
Approach
How I work
-
Architecture you can ship
I own decisions end-to-end, from model and framework choice to deployment, security and performance.
-
Hands-on, not hand-off
I write code, build the frameworks teams use and review the code that ships.
-
Secure by design
OAuth 2.0, JWT, encryption and compliance are part of the first design, not a later audit.
-
Teams that grow
Agile delivery with mentoring and coaching, so teams keep improving after the project ends.
Range
Industries I've built for
- AI startups
- Public sector
- Aviation & logistics
- Telecom
- Retail
- Blockchain & Web3
- Legal tech
- Pharma
- AgTech
- Gaming
- Cybersecurity
- Digital media
- Consumer electronics
Capabilities
Skills, with proof
Grouped by what I deliver, with links to the work where each was used.
-
Agentic AI and LLM apps
Agents that plan, call tools and act on real systems, with memory and guardrails.
- AI agents
- Multi-agent systems
- Agent orchestration
- Agentic memory
- Function calling
- MCP
- LangGraph
- LangChain
- CrewAI
- Prompt engineering
-
RAG and AI evaluation
Retrieval that grounds answers in your data, measured so quality can be tracked.
- Retrieval-augmented generation
- Semantic search
- Information retrieval
- LlamaIndex
- pgvector
- Neo4j
- RAGAS
- AI evaluation
-
Computer vision and ML
Models trained, evaluated and shipped into production apps, from detection to prediction.
- Computer vision
- YOLO
- ResNet
- PyTorch
- TensorFlow
- NLP
- Predictive analytics
- Fine-tuning
-
Cloud-native architecture
Event-driven, serverless platforms on AWS that stay secure and scale with load.
- Solution architecture
- Event-driven architecture
- Microservices
- AWS Lambda
- API Gateway
- Amazon Bedrock
- Docker
- CI/CD
- OAuth 2.0
-
Full-stack and APIs
Hands-on delivery across backend, frontend and mobile, from schema to UI.
- Python
- FastAPI
- Django
- SQLAlchemy
- Node.js
- TypeScript
- React
- Flutter
- PostgreSQL
- MongoDB
-
Leadership and delivery
Turning business goals into architecture, then leading teams to ship it.
- Technical leadership
- Architecture governance
- Mentoring
- Code review
- Requirements gathering
- Agile delivery
- GenAI strategy
- Consulting
Experience
Where I've worked
-
Mar 2025 – Present 1 yr 8 mo Now
Agentic AI Architect
Independent consultant
Agents, RAG and vision systems for SaaS and AgTech ventures
- Document AI
- Computer vision
- LLM + RAG apps
- MCP servers
- LangChain
- LangGraph
- AWS Bedrock
- OpenRouter
- MCP
- YOLO
- ResNet
- +4
Key highlights
Fractional technical architecture and hands-on AI/ML engineering for early-stage SaaS and AgTech ventures. Embedded with each client's team to scope, architect, build and ship production AI systems end-to-end, using AI-assisted development and multiple LLM providers.
- Document AI: OCR extraction plus LLM classification and schema-conformant JSON output via AWS Bedrock and OpenRouter, deployed as isolated per-client AWS accounts
- Computer vision: two-stage pipeline with YOLO object detection and ResNet classification for disease, weight and breed, trained on IoT and CCTV data and shipped to mobile and web apps
- LLM and RAG app: LangChain pipelines with tool calling, conversational memory and RAG over a predictions knowledge base, translated natively into five languages
- Built a Model Context Protocol (MCP) server exposing coupon issuance, redemption and merchant operations to LLM agents
-
Dec 2023 – Feb 2025 1 yr 3 mo
Technical Architect Team lead
Large IT services provider, Singapore
Public sector and aviation platforms
- Microservices migration
- Serverless on AWS
- Government-grade security
- LLM proofs of concept
- Go
- Python
- FastAPI
- NestJS
- GraphQL
- AWS Lambda
- ECS Fargate
- +9
Key highlights
Architected and delivered compliance-driven platforms for public sector and aviation logistics, owning architecture end-to-end including AI/LLM initiatives.
- Led monolith-to-microservices migrations; built serverless frameworks in NestJS, Go and Python serving GraphQL and REST on AWS Lambda and API Gateway
- Containerised services on Docker and AWS ECS Fargate to cut infrastructure overhead
- OAuth 2.0 and JWT authentication aligned with government security guidelines; AES-256 and SHA-3 for data protection
- Decoupled real-time messaging with Redis Pub/Sub and Amazon SNS topic subscriptions
- Built AI/LLM proofs of concept with Gemini, OpenAI and Ollama for citizen services and workflow automation
- Delivered a multi-agency citizen data platform (Go, MongoDB, DynamoDB, PostgreSQL), an aviation data hub (Python ETL on Databricks, AWS Glue, Kinesis) and an AI cargo sequencing system (FastAPI, Strawberry GraphQL, SQLAlchemy)
- GitLab CI/CD, load and performance testing against public-sector SLAs, structured logging and observability
-
Oct 2019 – Dec 2023 4 yrs 3 mo
Solution Architect Team lead Blockchain
Global SaaS company, Singapore
Multi-tenant SaaS and blockchain coupon platforms
- Event-driven SaaS
- Blockchain smart contracts
- ERC-20 token sale
- Flutter mobile
- Ethereum
- Solidity
- OpenZeppelin
- Node.js
- Go
- Java Spring Boot
- Flutter
- +7
Key highlights
Owned the technical architecture of a global SaaS digital coupon platform supporting multi-tenant business models and high transaction volumes, and of its blockchain-based counterpart for coupon ownership and traceability.
- Blockchain coupon platform on Ethereum: Solidity smart contracts for B2C, B2B and C2C coupon creation, distribution, transfer and redemption, with on-chain ownership tracking to prevent counterfeiting
- ERC-20 utility token and token-sale contracts built on OpenZeppelin: pricing, sale windows, hard cap, contribution limits, whitelist, allocation and claiming, vesting and lock-ups, refunds, admin roles and emergency pause
- Converted a monolith into event-driven, serverless microservices on API Gateway and Lambda in Node.js, Go and Java Spring Boot
- Set frontend standards (Vue + Nuxt, React + Next) after evaluating Angular, Vue and React
- Designed a Flutter mobile framework using Domain-Driven Design for iOS and Android
- OAuth 2.0, JWT and TOTP authentication; payment gateway test beds with SQS FIFO and standard queues
- Dynamic per-business subdomains, ECS Fargate deployments, Kafka and ELK for event processing and logs
-
Jul 2018 – Sep 2019 1 yr 3 mo
Technical Architect Team lead
Government technology agency, Singapore
Government procurement and approval systems
- Multi-level approvals
- Low-code delivery
- Audit and compliance
- OutSystems
- Low-code
- Oracle
- Private cloud
- Agile/Scrum
Key highlights
Led delivery of a next-generation procurement platform covering purchase requests and orders, sourcing, tendering, contract management and award approval.
- Multi-level, multi-approver workflows with full audit trails on a low-code platform
- Data modelling and integration design on the OutSystems low-code platform with Oracle, on a private cloud
- Led Agile/Scrum teams, code reviews and security and compliance standards
-
Jan 2017 – Jun 2018 1 yr 6 mo
Technical Architect Team lead
Telecommunications provider, Singapore
Telecom account planning and task automation
- Sales task automation
- REST integrations
- Data migration
- OutSystems
- Low-code
- REST APIs
- Microservices
- Oracle
Key highlights
Led solution architecture for an enterprise account-planning system that automatically creates, assigns and approves periodic sales tasks across a role hierarchy.
- Translated business processes into application and technical architecture components
- Built RESTful services and third-party API integrations on the OutSystems low-code platform
- Planned data migration and rollout, and led peer reviews and quality management
-
Mar 2013 – Dec 2016 3 yrs 10 mo
Technical Solution Architect Team lead
Software consultancy, Singapore
SaaS products for clients in the UK, Europe and Asia
- Document automation SaaS
- AI chatbot SaaS
- Multi-tenant data
- Payments and SSO
- React
- Next.js
- Vue.js
- Node.js
- Java Spring Boot
- AWS Lambda
- Azure Form Recognizer
- +3
Key highlights
End-to-end design and delivery of frontend and backend systems for client SaaS products: module division, interface design, data flow, technology selection and code.
- Legal invoice-management SaaS with automated invoice scanning using Microsoft Form Recognizer (React, Next.js, Node.js, AWS Lambda)
- Customer-support SaaS with an AI chatbot, tickets and appointments (Vue, Nuxt, Go, Flutter apps)
- Sales-force automation SaaS for pharmaceutical companies (Java Spring Boot, iOS and Android apps)
- Multi-tenant schema design across Oracle, MySQL, MongoDB, DynamoDB and CouchDB, plus Neo4j graph modelling
- OAuth 2.0, JWT and third-party sign-in (Azure AD, Google, Facebook); regional and international payment gateways
Earlier career
-
Technical Lead
Retail voucher and store back-office systems · Retail group, Singapore
From Mar 2009
-
Technical Lead
Signature-less antivirus engine · Cybersecurity software company, Singapore
From May 2006
-
Senior Software Engineer
UPnP multimedia streaming server · Digital media company, India
From Oct 2004
-
Senior Software Engineer
Embedded media players and Windows device drivers · Embedded systems company, India
From Mar 2000
Toolbox
Technology stack
AI/ML Frameworks
- Python
- TensorFlow
- PyTorch
- Scikit-learn
- Keras
- YOLO
- ResNet
GenAI & Agentic Orchestration
- LangChain
- LangGraph
- PydanticAI
- MCP
- CrewAI
- Semantic Kernel
- LlamaIndex
- Haystack
LLM Models & Platforms
- OpenAI GPT
- Anthropic Claude
- Google Gemini
- Azure OpenAI
- Microsoft Foundry
- AWS Bedrock
- OpenRouter
- Meta LLaMA
- Mistral
- Ollama
AI-Assisted Development
- Claude Code
- Cursor
- Google Antigravity
Vector Databases & Knowledge Graphs
- Pinecone
- ChromaDB
- FAISS
- PostgreSQL (pgvector)
- Neo4j
Document AI & Computer Vision
- OCR
- Intelligent Document Processing
- Computer Vision Pipelines
MLOps & Deployment
- MLflow
- Docker
- Kubernetes
- Azure ML
Automation & Integration
- n8n
- Azure Logic Apps
- REST APIs
- Webhooks
- CRM/ERP/HR/BI Integration
Cloud Platforms
- AWS (Lambda, API Gateway, ECS Fargate, EC2, S3, RDS, DynamoDB, SNS, SQS, WAF, CloudFormation)
- GCP
- Azure AI/ML
- Azure DevOps
- Microsoft Fabric
Show 9 more technology groups
Backend & APIs
- Python
- FastAPI
- Node.js
- NestJS
- Go
- Java (Spring Boot)
- C#
- GraphQL
- Django
- Express
- SQLAlchemy
Blockchain & Web3
- Ethereum
- Solidity
- ERC-20
- OpenZeppelin
- Etherscan
Mobile & Low-Code
- Flutter
- iOS
- Android
- OutSystems
Data Engineering
- Databricks
- PySpark
- AWS Glue
- Kinesis
- ETL pipelines
Frontend
- React
- Next.js
- TypeScript
- Vue
- Nuxt
Databases
- PostgreSQL
- MongoDB
- DynamoDB
- MySQL
- Oracle
Messaging & Caching
- Kafka
- Redis
- Amazon SNS/SQS
DevOps & Platform
- Docker
- Kubernetes
- GitLab CI/CD
- ECS Fargate
- Observability (ELK)
Security
- OAuth 2.0
- JWT
- AES-256
- IAM
- API security
- AWS WAF
Education
Education & certification
- Master of Science, Artificial Intelligence & Machine Learning, Liverpool John Moores University, UK
- Advanced Certification, Machine Learning & Cloud, Indian Institute of Technology (IIT), India
- Master of Computer Applications (MCA), Bharathidasan University, India
- Bachelor of Science, Computer Science, Bharathidasan University, India
Have an AI system to design or ship?
Tell me about the problem, the data you have and your timeline. I reply to every message.