Intelligence,
engineered
to work.
I’m Vivek. I build AI-powered SaaS, full stack applications and agentic systems that turn complex workflows into useful products.
Thoughtful product.
Intelligent underneath.
Systems with
something to solve.
A selection of agent workflows, product platforms and the infrastructure behind them.
Descriptive project names. Interface studies illustrate the system concepts.
From first idea.
To the real world.
One engineering partner across the model, the product and the systems that keep it running.
Agents that do the work
Stateful agents, connected tools and durable workflows. Retries, human approvals and useful context at each step.
LangGraph / Temporal / MCP · A2A / RAGProducts people can use
AI-powered SaaS, full stack applications and integrations. Thoughtful interfaces with dependable backend services.
React / Next.js / Python / GoModels that fit the task
LoRA fine-tuning, evaluation harnesses and local inference. Improve model behavior with curated data and measurable checks.
Hugging Face / LoRA / Ollama / vLLMData that drives decisions
SQL analytics, batch processing and event-driven data pipelines. Scheduled with clear dependencies, data-quality checks and model refreshes.
SQL / Spark / Snowflake / Airflow / KafkaThe right tools.
For the right problem.
Agentic AI & automation
AI agents that connect your tools, knowledge and people. From RAG assistants to durable multi-agent workflows with human oversight.
Explore the service ↗Full stack & AI-powered SaaS
Web applications and SaaS products with clear interfaces, dependable APIs and AI features that serve a real business need.
Explore the service ↗Data analytics & model engineering
Turn operational data into usable insight and model-ready datasets. Connect analytics, fine-tuning and evaluation to the application.
Explore the service ↗Backend, cloud & Web3
The systems behind the product: APIs, event streams, cloud services and wallet-connected applications.
Explore the service ↗Messaging bots & commerce automation
WhatsApp bots, Telegram automation, SMS workflows and Shopify integrations that connect conversations to your business operations.
Explore the service ↗Good software
is a team sport.
Merged contributions to the tools developers use every day. Real pull requests, open for you to explore.
My GitHub
THE HUMAN BEHIND THE SYSTEMS ↗Curious by default.
Engineer by practice.
I’m Vivek Singh, an AI systems and full stack engineer with four years of experience across products, microservices and agent workflows.
I like the parts where disciplines meet: an agent that needs the right context, a backend that needs to stay reliable, an interface that makes the whole thing feel simple.
Practice, backed
by perspective.
Security fundamentals, alongside recognition for hands-on engineering and problem solving.
Security, Compliance &
Identity Fundamentals
View credential ↗Tech-A-Thon · 2023
Google Hash Code · rank 2,740
A few useful
answers.
India based. Remote collaboration across India and internationally, with working hours agreed for each project.
What can I hire you to build?
Agentic AI systems, AI-powered SaaS, full stack websites and applications, APIs, WhatsApp/Telegram bots, Shopify integrations, data analytics and cloud platforms. I also work on model fine-tuning, evaluation and Web3 integrations when they fit the project.
Are you available for full-time and contract work?
Yes. I am based in India and open to freelance, project-based and longer-term remote contracts, as well as suitable full-time engineering roles. We can agree on hourly, weekly, monthly or milestone-based arrangements after reviewing the scope.
Do you work with teams across India and internationally?
Yes. I collaborate remotely with teams across India and international markets, including Europe, Canada, the UK and the UAE. We agree on time-zone overlap, communication and delivery expectations at the start.
Can you connect agents to our existing tools and data?
Yes. The work can include REST or GraphQL APIs, MCP tools, web search, retrieval over internal knowledge, workflow orchestration and human review. Access and permissions are scoped to the integration.
Do you handle model training and improvement?
I work with LoRA fine-tuning, curated datasets, evaluation harnesses and local inference using tools such as Hugging Face, Ollama and vLLM. We choose between prompting, retrieval and fine-tuning based on the task, then evaluate changes before release.
How do we start?
Email the business problem, your current stack and any deadline or constraints. We can define a bounded first milestone, acceptance checks and a negotiable payment structure.
Have something
worth building?
Tell me what you’re working on.
Let’s find a clear first milestone.