Software Engineer II @ JPMorgan Chase · Glasgow, UK

Hi, I'm Mohammed Kaish Ansari

I build

Software Engineer with 5+ years designing high-performance big-data systems and taking AI capabilities from experiment to production in fintech — from petabyte-scale Spark pipelines to enterprise agentic tooling powered by the Model Context Protocol.

rocket_launch View my work description Résumé
Mohammed Kaish Ansari
bolt 550M+ records
smart_toy AI & LLM Systems
0
Years of experience
0
MCP servers integrated
0
Records processed
0
Cloud cost reduction

About me

Engineering reliable systems at data & AI scale

I'm a Software Engineer II on the Liquidity Risk, Big Data team at JPMorgan Chase in Glasgow, where I design and maintain the large-scale Spark and Snowflake pipelines behind daily regulatory stress calculations reported to the Federal Reserve.

Lately my focus has been applied AI & developer productivity — architecting enterprise agentic tooling on top of the Model Context Protocol, building custom MCP servers, and establishing best practices for AI observability, reliability, and context engineering so that GenAI capabilities ship to production, not just demos.

I started out at Airtel Digital building real-time fraud-prevention and reconciliation systems over hundreds of millions of records, and I still love the craft of squeezing latency and cost out of distributed data systems. Competitive programmer at heart.

work
Current role
Software Engineer II · JPMorgan Chase
location_on
Based in
Glasgow, United Kingdom
school
Education
B.Tech, CSE · IIIT Una (2017–2021)
mail
code
Focus
Big Data · Cloud · AI & LLM Systems

Toolbox

Skills & technologies

The languages, platforms, and frameworks I reach for to build production systems.

terminal Languages

ScalaPythonC++ JavaTypeScriptGoSQL

smart_toy AI & LLMs

LLM IntegrationAgentic SystemsRAG Prompt & Context EngineeringLangChainLangGraph Pydantic AIOpenAI SDKMCPGitHub Copilot

cloud Cloud & DevOps

AWS EMR ServerlessEC2S3Athena SQSRDSIAMDocker KubernetesTerraformPrometheusGrafana

database Big Data & Pipelines

Apache SparkKafkaAirflowHadoop HiveSqoopSnowflakedbtSolace

storage Databases

PostgresMySQLOracleDynamoDB RedisHBaseImpala

security Auth & APIs

REST APIOAuth 2.0SAML SSO ADFSROPCKerberosNode.js

Career

Experience

5+ years across fintech and telecom, from real-time systems to enterprise AI tooling.

Software Engineer II — Liquidity Risk, Big Data
Feb 2024 — Present
JPMorgan Chase & Co. · Glasgow, UK
  • Enterprise AI tooling: architected a VS Code extension integrating 11+ MCP servers (Bitbucket, Confluence, Jira, Swagger, CI, Databricks) as GitHub Copilot tools, enabling LLM-driven agentic workflows for engineers.
  • Built custom MCP servers in TypeScript/Node.js — Knowledge (semantic search across 100+ doc repos), Terminal (guard-railed command execution), Skynet (EMR Serverless & S3 log analysis) and Memory (persistent cross-conversation context).
  • AWS EMR Serverless migration: led migration of 15+ Spark pipelines off EC2 EMR, cutting cost 30–40% and removing cluster-management overhead, validated with parallel testing across 1M+ records.
  • Performance: optimized stress calculators for 20% faster daily runs and 25% better query performance, cutting ETL runtime from 120 to under 90 minutes for timely Federal Reserve reporting.
  • Implemented a unified ADFS auth layer (SAML SSO, ROPC, cookie flows) and an automated OAuth 2.0 EMR cluster-management system reducing provisioning time by 60%.
Software Engineer — Postpaid, Big Data
Jul 2021 — Feb 2024
Airtel Digital (Bharti Airtel Ltd.) · Gurugram, IN
  • Real-time fraud prevention & dedupe engine: processed 550M+ customer records using Spark for batch and Kafka for streaming ingestion, with highly concurrent Go REST APIs delivering sub-second fraud-detection responses.
  • Data reconciliation system: built end-to-end integrity checks with Spark, Go, Python and Airflow, plus Telegram-bot alerting that cut mean-time-to-resolution by 40%.
Software Engineering & Research Internships
2019 — 2021
Bumblebee Data Inc. · Applied Science Lab, IIT Mandi
  • Bumblebee Data: developed Python/Spark programs in an Agile team and identified ML integration opportunities in the product (NLP, RAPIDS).
  • IIT Mandi (FarmerZone): built real-time data-cleaning and outlier-detection pipelines under Dr. Srikant Srinivasan.

Selected work

Projects & initiatives

Systems I've designed and shipped — from enterprise AI tooling to large-scale data platforms.

smart_toy

Enterprise Agentic VS Code Extension

JPMorgan Chase · Applied AI

A VS Code extension exposing 11+ internal MCP servers as GitHub Copilot tools, enabling LLM-driven agentic automation for enterprise engineers — with structured evaluation frameworks for reasoning quality and output reliability.

TypeScriptNode.jsMCPLLM
hub

Custom MCP Server Suite

JPMorgan Chase · Platform

Knowledge MCP for semantic search across 100+ doc repositories, Terminal MCP for guard-railed stateful command execution, Skynet MCP for EMR Serverless monitoring, and Memory MCP for persistent cross-conversation context.

MCPRAGAWSContext Eng.
savings

EMR Serverless Migration

JPMorgan Chase · Data Platform

Migrated 15+ Spark stress-analytics pipelines from EC2 EMR to AWS EMR Serverless, achieving 30–40% cost reduction and validating data parity across 1M+ records with a production parallel-testing strategy.

AWS EMRSparkSnowflakeScala
shield

Real-Time Fraud Prevention Engine

Airtel Digital · Scale

A fraud-prevention and deduplication engine over 550M+ customer records, combining Spark batch processing, Kafka streaming ingestion, and concurrent Go REST APIs with sub-second detection responses.

GoKafkaSparkREST
rule

Data Reconciliation System

Airtel Digital · Reliability

End-to-end data-integrity checking orchestrated with Airflow across Spark, Go and Python, with a Telegram-bot notification system that reduced mean-time-to-resolution by 40%.

AirflowPythonGoSpark
code

Open-Source & Early Projects

Personal · GitHub

A collection of deep-learning and tooling projects — YOLOv2 real-time text detection, LSTM stock-price prediction, and CoolKit (a Codeforces command-line contest tool) — from my university years.

PythonDeep LearningCLI
Browse on GitHub north_east
open_in_new See more on GitHub

Recognition

Honors & achievements

trophy

Systematic Trading Competition — 3rd Place

JPMorgan Chase & Co. · Nov 2025

Placed 3rd in the firm-wide systematic trading competition.

workspace_premium

Quantitative Research Virtual Experience

JPMorgan Chase & Co. · Jun 2025

Completed JPMorgan's Quantitative Research virtual experience program.

Get in touch

Let's build something reliable.

Open to conversations about data engineering, applied AI, and interesting problems at scale.

mail Say hello group Connect on LinkedIn