AI agents, RAG and automation
Build document search and Q&A with sources, connect your business tools, or repair a failing integration. Work directly with Rajiv on AI agents, RAG, AWS engineering and n8n automation.
Delivery view
Inputs
Documents Business data Customer eventsIntelligence
AI agents + RAG Rules + approvals MonitoringOutcomes
Answers with sources Actions in tools Operational insightWhat we do
Focused consulting and implementation for organizations that need AI connected to real data, workflows and operating controls.
Design and delivery for Bedrock agents, serverless application layers, contact centres and analytics workloads.
Document retrieval, semantic search and Q&A with sources, using OpenAI, pgvector or Amazon Bedrock to suit your data and requirements.
Reliable workflows across CRM, email, voice, customer support and internal operations.
Selected work
Explore paid client work and separate portfolio demonstrations. Each card states its scope; a demo is not evidence of production delivery.
Paid client work · Bedrock operations
Built for Emagination Cloud, the agent queries Parquet data through Athena and investigates failures from S3 debug logs. The client owned the frontend and production deployment.
Watch the walkthroughPaid client work · Infrastructure review
Reviewed 13 Terraform files and a Dockerfile before the client's first deployment. The paid two-hour engagement covered correctness, security, cost and first-apply risks.
Delivered written findings, including a migration script missing from the runtime image. The engagement was a review only, with no Terraform plan, code changes or deployment.
Discuss an infrastructure review
Synthetic-data POC · Healthcare RAG
Portfolio proof of concept combining React, cited document answers and Athena Text-to-SQL over synthetic healthcare data. No production healthcare or compliance claim.
Watch the walkthrough
Portfolio POC · Amazon Connect
Conversational Lex IVR, member authentication, context-aware handoff, Amazon Q agent assist and repeatable deployment.
Watch the walkthrough
Portfolio demo · QuickSight and Redshift
A governed analytics setup that gives each audience the right view while keeping the underlying warehouse centralized.
Watch the walkthrough
Portfolio build · Document Q&A
FastAPI RAG application for finding relevant internal knowledge and returning answers with visible supporting sources.
Read the project walkthrough
Portfolio demo · n8n automation
Claude extracts enquiries, logs structured leads, converts natural-language timing and routes uncertain results for human review.
Watch the walkthroughSynthetic-data POC · Trello and Slack
AI extracts enquiry details, creates a Trello card and sends a Slack notification, with validation and manual-review routing. A portfolio POC using synthetic data.
Watch the demoClient proof
Client feedback across AI agents, RAG, automation and infrastructure reviews. Portfolio demonstrations are labelled separately.
Verify on UpworkAWS Bedrock and Terraform · Emagination Cloud
“Everything looks GREAT! The AWS components operate exactly as expected and the updates to the Terraform files are perfect I'm super happy with both the quality and turnaround time.”
Terraform pre-deployment review
“Rajiv reviewed a Terraform configuration that had never been applied, before I ran it for the first time. The report was super precise, verifiable and more helpful than I expected. Every finding pointed at a file and a line, with the reasoning behind it rather than a tool dump, which is what I had asked for. So a very very helpful session and highly recommended”
WhatsApp AI Customer Support Agent
“Rajiv built us an AI support agent for WhatsApp that handles our repetitive customer questions automatically: order status, returns, shipping times, product info, opening hours. It uses RAG to pull answers from our knowledge base, so responses are accurate and match our actual policies, not generic replies. When a query is too complex, it hands off to a human agent with full context, and every conversation gets logged in Google Sheets. The whole thing runs through n8n connected to WhatsApp Business API, and we can add new FAQs ourselves without needing a developer. Rajiv walked us through the setup clearly, and communication was fast throughout. Highly recommend him for n8n, WhatsApp automation, RAG, and customer support AI projects.”
AI Automation Expert for Agency
“Rajiv is honestly super talented, I love working with him and he is a expert in AI automation and web development I can’t wait to use him again, couldn’t recommend enough. Also a super great guy in general!”
AI Voice Receptionist for Missed Calls
“Rajiv delivered exactly what we needed. We were losing leads every time we missed a call during busy hours and after closing. Now, an ElevenLabs voice agent picks up via VAPI, sounds completely natural, answers questions about our services and pricing, and books the caller straight into Google Calendar without us lifting a finger. What impressed us most was how natural the call flow feels; callers actually stay on the line and book. Every call ends with an instant Twilio SMS to my phone with the caller's name, what they needed, and whether they booked or left a message. The whole workflow runs on n8n, connected to our Twilio number and tested with real calls. Rajiv walked us through updating the script ourselves without touching any code. Will definitely hire again.”
AI Video Ad Generator with RAG and AWS
“Rajiv built us an AI automation system in n8n that turns a product idea into a finished AI video. It uses RAG to pull from our brand documents, writes the ad script, generates the AI video, and saves everything to AWS S3 with version control. He had already built similar AI video automation workflows using RAG and n8n, which gave us confidence in his expertise. The final solution matched our requirements perfectly, and I highly recommend him for AI video generation, RAG, n8n, and AI automation projects.”
AI Content Generation and Auto-Publishing
“Rajiv is a pleasure to work with. He demonstrated good technical expertise in building our AI content generation and auto-publishing system using n8n and AI models. His knowledge of automation, AI workflows, and n8n was evident throughout the project. I'm very happy with the quality of his work and would highly recommend him to anyone.”
How engagements work
Map the workflow, data, users, constraints and success measure.
Choose the smallest design that meets security and operating needs.
Build, integrate and test against realistic cases and failure paths.
Document ownership, monitoring, costs and the next sensible improvement.
About the founder
Rajiv Unnikrishnan brings 18+ years in enterprise technology across AWS and Oracle, including partnerships and business roles. FlowboticsAI is his independent freelance business, focused on AI agents, RAG, AWS engineering and workflow automation.
You work directly with Rajiv. Start with a scoped paid diagnostic, a written infrastructure review or an implementation milestone, with delivery responsibilities agreed upfront.
Start with the workflow
Use a 30-minute discovery call to discuss your data, systems and delivery needs. We can then scope a paid diagnostic, review or implementation milestone.
