Skip to content

Mahesh Kandula

AI Engineer • Coventry, United Kingdom • linkedin.com/••••• • drivetube.ai/•••••

Professional Summary

AI Engineer with 2 years of experience designing and deploying intelligent document processing, NLP, and generative AI solutions in regulated financial-services environments. Specialises in building LangChain-based RAG pipelines, fine-tuning open-source LLMs using LoRA and QLoRA, and applying BERT-based NER and OCR extraction to automate compliance and fraud-detection workflows at American Express. Comfortable across the full ML lifecycle — from feature engineering and model training with XGBoost and Hugging Face Transformers to containerised deployment on AWS SageMaker using Docker, MLflow, and Jenkins CI/CD. Brings additional exposure to computer vision, PySpark-based data processing, and responsible-AI documentation practices. Currently deepening expertise through an MSc in Data Science and Computational Intelligence at Coventry University.

Technical Skills

Programming Language: Python
Cloud Platforms: AWS,GCP
DevOps & Infrastructure: Docker,Jenkins,GitLab CI
Data Engineering & Processing: PySpark,Apache Airflow
Machine Learning & AI: Gradient Boosting,Named Entity Recognition,Document Classification,OCR
AI/ML Frameworks & Libraries: scikit-learn,XGBoost,Hugging Face Transformers,BERT
Generative AI & LLMs: LangChain,LangSmith,LoRA,Fine-Tuning,QLoRA,Llama,Mistral
Vector Databases & RAG: Pinecone,FAISS,Retrieval-Augmented Generation

Work Experience

American Express
India
AI Engineer
Aug 2023 – Aug 2025
Worked within American Express's AI/ML practice supporting fraud-detection and document-intelligence platforms across financial-services compliance and customer-operations workflows.
Tech Stack: LangChain, LangSmith, Retrieval-Augmented Generation, Pinecone, FAISS, BERT, OCR, XGBoost, Gradient Boosting, Docker, MLflow, AWS SageMaker, Jenkins, GitLab CI
  • Architected and deployed Retrieval-Augmented Generation pipelines using LangChain over compliance and customer document corpora, applying LangSmith-based evaluation to measurably improve LLM response relevance and citation accuracy.
  • Built document classification and information-extraction components for the Intelligent Document Processing platform, processing PDFs and scanned compliance documents at scale to reduce manual review effort by over 60%.
  • Applied BERT-based Named Entity Recognition and OCR extraction to identify and surface structured fields — counterparty names, transaction amounts, and policy identifiers — from unstructured financial documents.
  • Trained and tuned XGBoost and Gradient Boosting models with rule-based logic to automate anomaly detection across document and transaction workflows, cutting false-positive alert volume by approximately 35%.
  • Containerised and shipped ML models to AWS SageMaker using Docker and MLflow, integrating Jenkins and GitLab CI pipelines to enable consistent, repeatable model deployments with automated regression checks.
  • Integrated Pinecone and FAISS vector databases into the RAG retrieval layer, benchmarking embedding strategies to optimise semantic search precision for compliance query workloads.
  • Backed responsible-AI reviews by authoring model-evaluation criteria and bias-mitigation control documentation, strengthening audit readiness across regulated financial-services products.
  • Delivered design and requirements sessions for fraud-detection AI workstreams, translating stakeholder compliance requirements into technical specifications for the ML engineering team.

Projects

Automatic Detection of Cardiac Arrhythmias from ECG Signals
Tools Used: Python, Scikit-learn, Signal Processing, Machine Learning, Classification Metrics
  • Developed a Python-based ML pipeline to automatically detect and classify multiple cardiac arrhythmia types from raw ECG signal data.
  • Engineered time-domain and frequency-domain features from ECG recordings and evaluated model performance using precision, recall, F1-score, and ROC-AUC metrics.
  • Benchmarked multiple classification algorithms to identify the most accurate and computationally efficient model configuration for real-time arrhythmia detection.
Enterprise Document-Intelligence RAG Assistant
Tools Used: LangChain, Pinecone, FAISS, GCP Vertex AI, NLP, LLMs, Python
  • Built a retrieval-augmented Q&A assistant over fund and account documents using LangChain, enabling compliance and support teams to surface cited, context-aware answers from large document repositories.
  • Integrated Pinecone and FAISS as the vector retrieval backend, reducing average document lookup time for compliance teams by significantly cutting manual search effort.
  • Deployed the assistant on GCP Vertex AI, managing embedding pipelines and model serving infrastructure to support multi-user concurrent query workloads.
Fine-Tuned Financial-Domain LLM
Tools Used: LoRA Fine-Tuning, QLoRA Fine-Tuning, Llama, Mistral, Python, Hugging Face Transformers
  • Applied LoRA and QLoRA parameter-efficient fine-tuning to open-source Llama and Mistral models on proprietary compliance and transaction datasets to improve document-understanding accuracy.
  • Reduced hallucinated outputs in domain-specific financial queries by evaluating fine-tuned model responses against a held-out compliance benchmark.
  • Optimised GPU memory utilisation during fine-tuning by selecting appropriate quantisation settings, enabling training on constrained compute budgets.
Fraud-Detection Virtual Assistant
Tools Used: LangChain, Vector Databases, Document Classification, Python, NLP
  • Designed an AI-powered virtual assistant that integrated document-classification logic and vector database retrieval to triage incoming fraud alerts automatically.
  • Reduced manual fraud investigation time by routing high-confidence alerts to automated resolution paths and surfacing relevant case history for analyst review.
  • Built the classification layer using fine-tuned NLP models capable of distinguishing fraud-signal document types across multiple product lines.

Education

Coventry University
MSc Data Science and Computational Intelligence • Coventry, United Kingdom • Sep 2025 – Aug 2026
Bharath Institute of Higher Education and Research
BTech Computer Science and Engineering • India • Jul 2019 – Aug 2023

Powered by Drivetube · Create your own profile at drivetube.ai

Everything on Drivetube

Six products, two of them free forever. Start wherever you are.

Read the docs →
  • Your free digital resume

    at drivetube.ai/in/your-name: one true standard resume with a Hiring Snapshot (visa status, expected salary, notice period, work preference, relocation), an ATS-ready PDF download and a single shareable link. Free forever; interview requests come from verified employers and your contact details stay masked until you accept.

    Documentation
  • Verified jobs, posted in the last 3 days

    verified openings crawled ATS-by-ATS from 140,000+ real company career pages across 50+ ATS platforms. Shows the true posting date from the source ATS — not when a listing was indexed — and deletes every general listing 3 days after it was actually posted. No ghost jobs, no ad-sponsored listings, no staffing reposts, no account needed.

    Documentation
  • Written by senior career writers

    human-written, ATS-optimised resumes by senior career writers, from ₹499.99 / $25.99. Available in every country, written to the destination country's own standard — a US resume, UK CV, German Lebenslauf and Indian resume are genuinely different documents. A paid service, separate from the free Drivetube Profile.

    Documentation
  • Drivetube OneFrom $39.99 / 3 mo

    Daily matched jobs, tailored resumes, or we apply for you

    Drivetube's paid service. Community sends daily Jobsheets of fresh jobs matched to your profile on WhatsApp and email, with the Drivetube Community ($39.99 for 3 months or $69.99 for 6 months; ₹999 or ₹1,899 in India). Jobhunt is one-time application credits with Community included: one credit tailors one resume and its cover letter (100 for $99 with 3 months of Community, 250 for $199 with 6 months). Let Us Apply is an add-on to Jobhunt: a human expert applies 250 times for you on your credits ($0.60 each, $150). Everything lives on one dashboard at drivetube.ai/drivetubeone.

    Documentation
  • For employers — no job postings, no applications

    hiring with no job postings and no applications. Paste your real job description and AI matches it against candidates' true standard resumes, returning ranked candidates with a match %, matched and missing skills and written reasoning. Free tier included; employers pay, candidates never do.

    Documentation
  • Documentation

    every product explained in full, with feature-by-feature comparisons against the job boards, AI apply tools, resume services and hiring platforms people actually use.

  • Playbooks

    tactical job-search guides. Each post is one named tactic with the situation it applies to, the exact steps, a copy-paste message and honest failure modes. Free, no account.

The job board covers the United States, India, United Kingdom, Canada, Europe and Australia, and Resume Writing Services are available in every country.

Pricing is identical on the web, iOS and Android — no app-store markup. Monthly Community membership is purchased in the iOS app only; the website sells annual plans. The Android app has no in-app purchase, so Android members subscribe on drivetube.ai and then sign in to the app with full access. Membership follows your account, not your device.