Srivathsan Suresh AI Engineer · Edinburgh, UK
Available for AI roles

Intelligent systems, from data to answers you trust.

See my work Get in touch
Scroll to explore
About

I'm Srivathsan — a consultant who builds production-grade AI applications. I turn ambiguous business problems into LLM & RAG systems that actually ship — grounded, evaluated and cost-aware, and built to survive real users, not just demos.

MScBusiness Analytics · Distinction
BEComputer Science · First Class
5×AWS & Oracle certifications
UKEdinburgh · open to relocate

Projects

Production LLM, RAG and data systems — designed, built and shipped.

AI Data Workspace

Meldify.Space

A production data workspace that turns raw spreadsheets into a clean, queryable dataset. A Neural Engine profiles the schema, detects structural and semantic quality issues and auto-repairs them; a natural-language Studio then answers analytical questions and generates full reports — built to handle hundreds of thousands of rows.

PythonFastAPILLM AgentsSchema ProfilingNeural Engine
preventive-care.appGrounded
What lifestyle changes lower my risk of type-2 diabetes?
Aim for ~150 min of activity a week, a fibre-rich diet, healthy weight and limiting refined sugar. Regular HbA1c screening helps catch early risk.
Grounded · NICE / NHS guidelines · not medical advice
What's the weather in London today?
That's outside my scope — I only cover preventive health, so I can't find an answer for that. Try asking about screening, nutrition or lifestyle.
Preventive-Care Assistant · Clinical RAG

Preventive Health Assistant

A preventive-medicine assistant answering patients on screening, nutrition, lifestyle and risk reduction — grounded with retrieval-augmented generation over clinical guidelines. Every answer is citation-backed; safety guardrails detect out-of-scope and emergency intents and refuse or escalate; PII is redacted and responses are evaluated against a clinician-reviewed golden set for factuality.

RAGGuardrailsClinical GuidelinesEval HarnessPII Redaction
retail-support.appRetrieving
There's a £240 payment to "QuickPay Ltd" I don't recognise.
Intent · possible fraud · £240 card txn
Could it be a subscription or a purchase under another trading name?
No — I've never used them.
Retrieving · card-dispute policy + prior cases
Is your card still with you — and have you shared your PIN or a one-time passcode with anyone?
Card's with me, and I've shared nothing.
Fraud signal elevated · card safety confirmed
I've blocked the card and opened dispute FD-4821. You're protected under PSR 2017 — a chargeback can take up to 15 working days, and I'll keep you updated.
Customer Support · Graph-RAG

Retail Banking Support Copilot

A customer-support assistant for everyday banking — cards, payments, transactions and fraud disputes — grounded on a knowledge graph built from policies and resolved cases. It classifies intent, extracts entities, detects fraud signals, retrieves and reranks the right sources, and returns a citation-backed answer with a confidence score — refusing when unsure and escalating securely to a human agent, with FCA-aware guardrails and full latency/cost monitoring.

Graph-RAGVector + Graph DBIntent / NERFraud signalsGuardrailsObservability
Capabilities

The stack I build production AI with.

Languages

01
PythonSQLJavaScript

AI & LLM Tooling

02
LangChainLangGraphMCPClaude CodeOpenAI CodexRAG

Backend & Web

03
FastAPINext.jsREST APIs

Data & Vectors

04
PostgreSQL / pgvectorOracle SQLWeaviatePinecone

DevOps & Infra

05
DockerKubernetesCI/CD

Cloud

06
AWSAzureGCP
Experience
May 2026 — Now

Senior Consultant — Wealth & Insurance

Lloyds Banking Group

Managing the claims process across the retirement journey — pension and investment decisions — conducting anti-money-laundering checks and applying FCA conduct rules.

Edinburgh, UK
May 2025 — May 2026

Software Engineer — Automations & Reporting

BriTek Technologies

Built an iOS app feeding a RAG pipeline (LangChain + PostgreSQL) that recommends LED replacements and generates AI-assisted proposals; shipped a full solar-PV sales platform and an e-commerce storefront end to end — with savings calculators, quote generation and a customer dashboard.

Manchester, UK
2024 — 2025

Services & Operations Coordinator

Menzies Aviation · Star Alliance

Managed passenger rebooking and disruption handling during irregular operations (IROPS) across multiple carriers using Amadeus Altéa DCS, under time-critical, safety-regulated conditions.

Edinburgh, UK
Jun 2021 — Sep 2023

Business Analytics Consultant

Oneglobe Systems · Client: PwC

Built PL/SQL pipelines and RPD data models in Oracle Analytics Cloud for PwC advisory, assurance and tax — powering SLA, engagement-margin and revenue reporting, and delivering KPI dashboards in OAC and Power BI.

Chennai, India
Feb 2020 — Jul 2020

Graduate Innovation Engineer

Forge Accelerator

Built a student-enrollment prediction model in Python (scikit-learn), a Power BI marketing-attribution dashboard, and behavioural cluster analysis to drive decisions.

Coimbatore, India
Credentials

Certifications

AWS Certified Data Engineer — AssociateAmazon Web Services · DEA-C01
View ↗
AWS Certified Cloud PractitionerAmazon Web Services · CLF-C02
View ↗
Oracle AI Vector Search ProfessionalOracle · 1Z0-184-25
Verify ↗
Oracle Enterprise Analytics ProfessionalOracle · 1Z0-1041-24
Verify ↗
Oracle APEX Cloud Developer — ProfessionalOracle · 1Z0-770
Verify ↗

Education

MSc Business Analytics — Distinction

University of Dundee — Scotland, UK
Graduated 2025

BE Computer Science — First Class

Kumaraguru College of Technology — India
Graduated 2021
Volunteering
Data Consultant — Homes For Good

Volunteering on a CRM / data consolidation project for a Glasgow social landlord supporting ~750 tenants — unifying fragmented data and reconfiguring Salesforce.

Contact

Let's build
something intelligent.

Open to AI / ML Engineer and Applied AI roles. If you're hiring or just want to talk shop about LLMs, RAG and data — my inbox is open.

Sriva6699@gmail.com