$Hello, World! I build data platforms and AI systems.

Jayanth Yanamandala

Software Engineer · Data Platforms & AI

Full-stack & platform engineer building data-platform and AI systems in fintech.

|New York, NYOpen to Work
Python
TypeScript
FastAPI
Flask
Kafka
SQS
Snowflake
PostgreSQL
LangGraph
RAG
AWS
Docker

// 01.about

About Me

A bit about who I am and what I do.

I build the systems that make data and AI usable in production — distributed pipelines, LLM gateways, retrieval infrastructure, and the observability around them. 3+ years of experience, mostly in fintech, where reliability and correctness aren't optional.

Currently at STEM Solutions LLC, building enterprise data-platform and AI systems for a leading financial-services firm: a distributed job-execution platform running thousands of pipelines, a firm-wide analytics service, and data-governance tooling — plus the RAG, multi-model gateway, and tool-using agent behind internal AI assistants, the access-control automation that provisions them, and the CI/CD and testing that ship it all. Previously at Visa, working on the critical path of real-time risk decisioning.

I hold an M.S. in Data Science from Stevens Institute of Technology. I'm open to Senior Software Engineer and AI Engineer roles — especially problems at the intersection of distributed systems, developer platforms, and production LLM infrastructure.

New York, NY

Education

M.S. Data Science

Stevens Institute of Technology

Jan 2023 – May 2024

B.Tech Computer Science & Engineering

JNTU Anantapur

2018 – 2022

3+

Years Experience

3

Companies

273K+

Pipeline Runs/Week

6K+

Pipelines on Platform

// 02.experience

Work Experience

Companies and teams I've shipped with.

STEM Solutions LLC

Software Engineer — Data Platform & AI

Current
STEM Solutions LLC·New York, NY
Feb 2025Present
  • Built a distributed job-execution platform running 6,000+ data pipelines (273K+ runs/week) for a leading financial-services firm: a scheduler dispatches to SQS and KEDA autoscales worker pods on queue depth for fault-isolated parallel execution, monitored via health checks and Grafana dashboards.
  • Designed an enterprise analytics service — the single source of truth for product usage — a pluggable instrumentation layer capturing every click and API mutation as session events, streamed via SQS → S3 → Snowpipe → Snowflake to power adoption and usage reporting across all apps.
PythonFlaskGraphQLVue.jsSQSKEDASnowflakePostgreSQLRAGLiteLLMAWS BedrockMCPTerraformPlaywrightVitestGrafana
Visa

Software Engineer — Risk Engineering

Visa·New York, NY
Aug 2024Feb 2025
  • Cut tail latency 30% on the critical path of a real-time risk-decisioning pipeline (Python, Kafka) with asyncio fan-out and request batching, while shipping 6 transaction-enrichment features into the authorization flow; hardened the service against cascading failures with circuit breakers, timeouts, and standardized error handling on the on-call rotation.
  • Built a rule versioning and rollback framework for a tier-1 fraud-decisioning service (200 configurable rules per transaction), letting Risk Strategy teams deploy and revert changes without engineering releases and cutting turnaround from several days to under 30 minutes.
PythonKafkaasyncioRule EngineCircuit BreakersPrometheusOpenTelemetryOpenAPI
Cognizant

Software Engineer

Cognizant·India
Jan 2022Dec 2022
  • Built and maintained 25+ REST APIs (Python, Django, PostgreSQL) for a Fortune 500 bank's loan-origination platform, cutting response latency 38% through query-plan optimization, N+1 elimination, and ORM-level caching.
  • Replaced a 12-hour nightly batch with a Kafka-backed reconciliation service — cutting data freshness from T+1 to under 5 minutes and reducing reconciliation tickets from 10/week to 2/week via automated mismatch detection and alerting.
PythonDjangoPostgreSQLKafkaRedis

// 03.projects

Projects & Open Source

Things I've built and shipped outside of work.

Active

Clausa

Agent-driven data observability — evidence-grounded root-cause analysis for data-quality incidents

A self-hostable data-observability platform where a LangGraph agent investigates warehouse anomalies and returns root-cause hypotheses grounded in verifiable evidence — profiling stats, lineage, pipeline runs, and schema changes. A verifier rejects any hypothesis not backed by real tool results, so the agent can't speculate its way to an answer.

2

LLM calls / investigation

4

Anomaly detectors

5

Read-only agent tools

168

Tests

PythonFastAPILangGraphLiteLLMPostgreSQLpgvectordbtSQSNext.jsTypeScriptDocker
Visit site

// 04.skills

Technical Skills

Technologies I work with regularly.

Languages

  • Python
  • TypeScript
  • JavaScript
  • SQL
  • Bash

Backend

  • Flask
  • FastAPI
  • Django
  • GraphQL
  • REST
  • asyncio
  • OAuth2 / JWT / RBAC
  • Microservices

Frontend

  • Vue.js
  • Vuetify
  • React 19
  • Next.js
  • Tailwind CSS
  • AG Grid

AI / ML & Agents

  • RAG
  • AI Agents
  • LangGraph
  • LiteLLM
  • MCP
  • pgvector / Milvus
  • Embeddings
  • RAGAS
  • AWS Bedrock
  • OpenAI / Anthropic / Gemini APIs
  • Multi-Agent Systems

Data & Cloud

  • PostgreSQL
  • Snowflake
  • Redis
  • Apache Kafka
  • Snowpipe
  • dbt
  • AWS (ECS, Lambda, S3, SQS)
  • Kubernetes / KEDA
  • Docker
  • Terraform

Testing & Observability

  • Pytest
  • Vitest
  • Playwright
  • GitHub Actions
  • Grafana
  • Prometheus
  • OpenTelemetry

// 05.contact

Get In Touch

Have a question, opportunity, or just want to say hi? I'll do my best to get back to you.

$ location: New York, NY
$ email: jayanthkyanamandala@gmail.com
$ response_time: within 48h