GANESH BOYA

• Sit with client teams early to understand how they actually work the workflows, the data gaps, the integration constraints and turn that into
  a technical plan the team can start building against within weeks, not months.
  FastAPI, SQL, and AWS code
  rather than just designing and handing it off.
  • Designed and deployed GenAI-powered engineering workflows using Devin, Claude Code, Cursor, and Anthropic Claude, accelerating
  software development, debugging, testing, code remediation, and developer productivity across enterprise environments.
  • Design agentic workflows that chain retrieval, internal tools, APIs, and a human checkpoint together, so multi-step business processes run
  safely without needing a person in the loop at every single step.
  • Connect AI applications into Snowflake, Databricks, internal databases, and enterprise SaaS tools so the agent works against the client's
  real data and systems, not a demo sandbox.
  • Build the retrieval pipeline from scratch ingestion, chunking, embeddings, vector search, reranking so responses stay grounded in the
  client's own documents instead of the model's best guess.
  • Put the controls in place that let a client actually trust an AI system in production authentication, governance, audit trails, and human-in-
  the-loop approval for anything higher-risk.
  •
  • Build interactive AI applications with React and TypeScript to help clients evaluate agent workflows and RAG experiences quickly before
  committing real engineering time, then turn the ones that land into systems the client can maintain long after I'm gone.
  • Write up the patterns and lessons from each engagement so the next one starts faster instead of rebuilding the same groundwork.
  • Build and ship agentic solutions on Amazon Bedrock Knowledge Bases for retrieval, Guardrails for safety, AgentCore and Strands Agents
  SDK for orchestrating multi-step, tool-using workflows.
  

Built production GenAI applications LLMs, RAG, tool-calling agents for enterprise use cases, taking each from a rough idea to something
a client team relied on daily.
• Developed Python/FastAPI services that wired foundation models into enterprise APIs, databases, and existing business applications so
the AI fit into workflows people already used.
• Designed retrieval systems document processing, embeddings, vector search, reranking, context construction to keep model answers
grounded instead of hallucinated, built AI agents that pull information across multiple sources, reason over enterprise context, and call
tools/APIs to carry out bounded business workflows on their own.
• Developed evaluation frameworks to catch accuracy and groundedness regressions early, so changes to models or prompts didn't silently
degrade what was already working in production.
• Put tracing, monitoring, authentication, and CI/CD in place to move AI applications from a working prototype to a service the team could
actually depend on. Partnered directly with enterprise stakeholders and engineering teams to turn business problems into technical
architectures they could actually deploy.
• Diagnosed production failures wherever they came from model behavior, retrieval, prompt/context issues, integrations and reworked the
system design until reliability held up, balanced solution quality against delivery speed and maintainability in fast-moving, often
ambiguous customer environments knowing when to ship and when to slow down.
• Applied MLOps/LLMOps practices prompt management, model monitoring to catch quality drift across model and prompt versions before
clients noticed it.

Built high-scale Python and Java services and data pipelines that powered real-time operational and product analytics.
• Designed distributed data-processing workflows for large volumes of event and transactional data, built to hold up under strict reliability
and fault-tolerance requirements.
• Developed REST APIs and backend services connecting internal apps, platform services, and data stores to their downstream consumers.
• Built batch and streaming pipelines in Python, Spark, Kafka, and SQL to handle ingestion, transformation, validation, and delivery of
production data, drove down pipeline latency and improved data quality, observability, and operational reliability across workloads.
• Optimized Spark and SQL processing paths and tracked down performance bottlenecks across distributed workflows.
• Worked closely with product managers, data scientists, and engineers to turn business requirements into scalable technical designs and
production systems.
• Owned services end-to-end development, testing, deployment, monitoring, and incident response including the root-cause analysis after
things broke.

Built backend services and data workflows supporting high-volume payment and financial-data processing in production.
• Developed Python and Java services, REST APIs, and SQL pipelines to move, transform, and serve transactional data reliably.
• Integrated backend applications with internal services, databases, and downstream systems while keeping data integrity and operational
• reliability front and center no small thing when the data is money.
• Worked with distributed systems where scalability, availability, and fault isolation weren't nice-to-haves they were the job.
• Improved monitoring, validation, and troubleshooting practices across business-critical payment workflows.
• Collaborated with product and engineering teams to turn business requirements into maintainable production solutions and unblock
delivery when things stalled.
• Applied solid engineering practices testing, code review, thoughtful API design across deployment, incident support, and performance
work