Lead AI Architect
I build multi-agent AI platforms, and RAG systems for real-world use.
Open to
Planner agent
LangGraphBreaks the request into steps and routes work to specialists
RAG retrieval
Azure AI Search · GraphRAG
MCP tool calls
JIRA · SharePoint · Web
Evaluation gate
Human-in-the-loopRAGAS scoring and an approval checkpoint before anything moves forward
Ship to production
Traces, alerts, runbooks, and named owners
From the blog
Enterprise Operating Model10 min read
Digital transformation strategies backed by data. See why the 70% failure stat is false, what BCG and Deloitte measured, and which metrics catch failure early.
Nvidia H1009 min read
Learn how to install Flash Attention 3 on NVIDIA Hopper GPUs, match CUDA and PyTorch, build from source, verify the package, and fix common errors.
Hugging Face Peft8 min read
Learn how to use QLoRA Hugging Face to fine-tune Llama 3 in 4-bit NF4 with PEFT and TRL, then compare VRAM use against LoRA and DoRA adapters.
Proof from the work
Banking, healthcare, energy, retail, recruitment, and enterprise platforms.
Selected work
What they have in common
Each case study starts with the user’s problem and shows the build, tradeoffs, and result.
Collaborator feedback
What comes through
They point to clear thinking, honest conversations, and a steady way of getting the work done.
“What I appreciated most about Swarnava was that he took the time to understand how our team works. We could ask questions without feeling we needed all the technical background first. He was easy to work with and thoughtful about the details.”
Emily CarterHead of Data & Analytics“Swarnava was someone I could talk ideas through with, even when they weren’t fully formed. He asked good questions and helped us work out what was worth pursuing. I also appreciated that he spoke up when something needed rethinking.”
Mei Lin ChenProduct Manager“I valued being able to hand something over to Swarnava and trust him to think it through. He was comfortable making decisions, but checked in when a trade-off needed a conversation. I’d be happy to work with him again.”
Daniel RossiCTO & Co-Founder“I had plenty of questions about where AI would actually be useful for us. Swarnava was patient and honest, including about its limitations. That helped us feel comfortable with the decisions we were making.”
Karen MüllerDirector of Business Intelligence“Swarnava worked well with the team from the start. He was open to feedback and took the time to explain his thinking during reviews. Even when priorities shifted, we could have a straightforward conversation and keep things moving.”
Thomas BennettVP of Engineering“In our conversations, Swarnava often caught details we’d overlooked. He asked about the information behind the answers, not just how convincing the output looked. I appreciated that care, especially when deciding what we could rely on.”
Ricardo AlvarezChief Data Officer“I didn’t always know the right technical questions to ask, and that was never a problem with Swarnava. He talked me through the options without making things overwhelming. I felt involved throughout, without having to oversee every detail.”
Aisha JohnsonFounder & CEOCareer so far
Across 6+ years
I’ve worked across cloud data platforms, semantic search, multilingual RAG, copilots, agent systems, and GraphRAG.
Leading an enterprise Knowledge Graph and GraphRAG platform for relationship-aware, grounded retrieval across enterprise knowledge
Driving the solution from discovery and architecture through production hardening with cross-functional teams
Cut intake-to-ticket turnaround from 2+ days to under 10 minutes, compressing P95 pipeline latency by 85% with parallelized Redis-backed workers
Improved learner skill-gap detection accuracy by 37% across 6K+ monthly sessions with adaptive quiz generation and continuous RAGAS evaluation
Scaled a supply chain AI copilot to 10K+ daily queries, reducing latency by 63% while maintaining 99.9% uptime
Boosted contextual recall by 34%, improved multi-turn response accuracy by 28%, and reduced context errors by 42%
Launched the company's first multilingual enterprise RAG platform and AI agent in production
Cut manual knowledge lookup by 43% and established BLEU and NDCG evaluation benchmarks
Built semantic search prototypes that drove adoption across 13 B2B users
Improved dataset usability by 26% for enterprise NLP work
Reduced nightly batch runtime by 95% for a global banking client
Decreased batch failures by 61% while preserving SLA compliance and improving ETL validation automation
Where I help
Design multi-agent systems with explicit tool access, memory boundaries, approvals, and specialist handoffs.
Combine semantic search with entity relationships when answers depend on both document meaning and connected business data.
Use fine-tuning, prompt design, dataset curation, and inference tuning to improve domain accuracy, latency, and consistency.
Design identity, data flow, deployment, monitoring, and ownership for AI systems that must run beyond the pilot.
About me
I'm Swarnava, an AI engineer and architect who takes multi-agent platforms, RAG, and GraphRAG from a clear business problem to a reliable production system.
More about meLet's talk
Best starting points
Share a consulting brief, full-time opportunity, architecture question, or AI system that needs a second opinion.
Calendar not loading? Open the booking page in a new tab.
Role or project
Share the role, problem, or timeline you have in mind. A short brief is enough.