I start with the workflow: what evidence enters, where a model helps, how its output is evaluated, and when a person should remain in control. This keeps RAG and AI agent work tied to a useful operational result.
Profile
Deepak Kumar Tiwari
Deepak Kumar Tiwari is a technology and AI engineering leader based in Udaipur, Rajasthan, India. Over 6+ years he has taken products from decision to operation — designing AI agents, RAG systems, and document intelligence, and building Golang microservices and cloud platforms that teams can run in production.
About
Where business context meets implementation
Deepak Kumar Tiwari works at the point where business context meets implementation. He leads technology delivery for software services organizations, translating requirements into architecture, delivery plans, and Statements of Work across healthcare, financial planning, enterprise reporting, HRTech, IoT, and digital media. He works hands-on rather than at a distance: reviewing code, shaping service boundaries, and keeping the operational path — cost, secrets, backups, observability, and rollback — in view from the start.
His applied-AI work stays tied to a useful operational result. That means designing retrieval and tool boundaries for RAG and AI agents, testing prompts against representative examples, making failure paths visible, and keeping human review where it matters. For privacy, latency, or cost-sensitive work, he evaluates smaller models and local inference — LoRA, MLX, Unsloth, and Llama.cpp — instead of assuming every workload belongs behind a hosted API.
- Full nameDeepak Kumar Tiwari
- RoleLead Software Developer · Technology & AI Leader
- LocationUdaipur, Rajasthan, India
- Experience6+ years in product, cloud, and AI engineering
- FocusAI agents, RAG, Golang microservices, AWS cloud architecture
- AvailabilityConsulting and delivery engagements, remote worldwide
Focus areas
What the work concentrates on
For privacy, latency, or cost-sensitive work, I evaluate smaller models and local inference before assuming every workload belongs behind a hosted API.
I define service boundaries, data ownership, deployment constraints, and operational paths so backend systems remain understandable as teams and traffic grow.
Working across frontend, mobile, APIs, and delivery planning helps me surface integration risk early and keep product decisions grounded in implementation.
Experience
Roles and outcomes
Leads technology delivery and client engagements across healthcare, financial planning, enterprise reporting, HRTech, IoT, and digital media.
- Translates business requirements into architecture, delivery plans, and Statements of Work.
- Guides stakeholder communication, engineering quality, and team mentoring across concurrent projects.
- Designs cloud-native, data-intensive, and applied-AI solutions with practical operational constraints.
Built and evolved production web, mobile, cloud, and data platforms for Indian and international clients.
- Designed distributed systems spanning 15+ cooperating services.
- Delivered data-governance, reporting, operational-dashboard, and real-time communication workflows.
- Improved delivery quality through architecture reviews, CI/CD, code review, and cross-functional planning.
Started in full-stack and mobile product delivery, building a foundation in JavaScript, React, Node.js, databases, and collaborative engineering.
- Contributed to multiple production web and mobile applications.
- Participated in code reviews, testing, releases, and customer-driven iteration.
Selected work
Systems that had to run
- AI-assisted engineering workflowAI
An open-source TypeScript CLI that turns project context and requirements into structured, repeatable engineering workflows.
- Custom healthcare analytics modulesData
Built custom analytics modules within a global healthcare enterprise’s chief data office program, connecting reporting needs to governed data and operational workflows.
- Hospital-at-home care platformHealthcare
Led end-to-end product delivery for hospital-at-home workflows spanning patient onboarding, care coordination, live monitoring, alerts, and operational visibility.
- Cancer-caregiving mobile experienceHealthcare
Delivered a mobile experience supporting cancer patients and caregivers through structured information, coordination, and role-aware support journeys.
- Live baseball streaming platformStreaming
Built product workflows for live baseball streaming, connecting media delivery, event context, audience experience, and operational controls.
- Distributed HRTech platformCloud
Designed service boundaries and delivery workflows for enterprise work management, documents, employee requests, and operational controls.
- Document intelligence systemsAI
Applied retrieval, agent workflows, multi-model orchestration, and local-model adaptation to turn unstructured documents into usable business knowledge.
- IoT learning platformLeadership
Delivered connected-device experiences for lesson playback, parental controls, progress visibility, and product operations.
Writing & contact
Keep reading, or get in touch
Common questions
Questions people ask first
Who is Deepak Kumar Tiwari?
Deepak Kumar Tiwari is a technology and AI engineering leader based in Udaipur, Rajasthan, India. He leads client delivery across healthcare, financial planning, enterprise reporting, HRTech, IoT, and digital media, and works hands-on with AI agents, retrieval-augmented generation, Golang microservices, and cloud architecture.
Where is Deepak Kumar Tiwari based?
He is based in Udaipur, Rajasthan, India, and works remotely with teams across India and internationally.
What does Deepak Kumar Tiwari specialize in?
Applied AI engineering (AI agents, RAG, document intelligence, local model adaptation), backend and cloud architecture (Golang microservices, AWS, Docker, PostgreSQL), and product engineering and delivery leadership (TypeScript, React, React Native, solution consulting).
How can I work with Deepak Kumar Tiwari?
Consulting, architecture, AI engineering, and senior technology leadership engagements usually start with a short conversation about your context and desired outcome. Details are on the contact page.