Alp – Python, Typescript, React

It's a personal software company focused on designing and building SaaS products, AI-powered workflows, and custom web platforms for real business use cases.

• Created tailored software solutions that delivered measurable business outcomes.
• Built RAG and ingestion/ETL pipelines for analytics and automation.
• Designed LLM orchestration and agentic workflows with evaluation and tracing.
• Shipped end-to-end web applications to production.

It's an AI-powered platform for generating DFMEA and PFMEA reports that turns technical documents into structured, industry-aligned failure analysis reports.

• Served as sole developer, owning end-to-end product from architecture to delivery.
• Designed system architecture with FastAPI back-end, Next.js App Router front-end, and PostgreSQL data layer.
• Built Gemini API-based AI pipeline, including model benchmarking for output quality and reliability.
• Used LangSmith to evaluate AI outputs and benchmark multiple models for quality, reliability and cost-performance.
• Implemented document processing and Excel export workflows for structured output generation.
• Managed GCP deployment with Terraform infrastructure and Keycloak authentication.
• Developed FMEA generation system using a reflection-agents approach for iterative review and refinement.

It's a SaaS platform that processes large volumes of U.S. Department of Labor Form 5500 filings and turns them into searchable, actionable business intelligence for financial advisors.

• Served as sole developer, designing and building a data-driven SaaS platform for financial advisors from scratch.
• Owned FastAPI back-end, Next.js App Router front-end, and PostgreSQL database design.
• Built Form 5500 ingestion and normalization pipelines to transform messy public filing data into structured product data.
• Developed search and filtering experience to surface plan-level opportunities and meaningful signals.
• Integrated Clerk for authentication.
• Managed Terraform infrastructure and GCP deployment.
• Shipped a fast, usable product that makes complex public filings actionable for advisors.

It's a data platform focused on the management, processing, and analysis of commercial property data, designed to handle high-volume data workflows and integrate reliably with external systems.

• Developed Python back-ends and data pipelines for the platform.
• Built scalable workflows for ingesting, validating, transforming, and synchronizing large volumes of commercial property data.
• Integrated the platform with external systems and third-party data sources for reliable data exchange.
• Designed and maintained back-end services, database logic, and asynchronous job flows for high-volume processing.
• Improved data quality and reliability through validation, error handling, logging, and monitoring.
• Contributed to Excel/VBA-based integration flows supporting operational users and business processes.
• Optimized PostgreSQL queries and data access patterns for performance and scalability.

It's a web application that simplifies compliance with EU medical regulations for laboratories and medical manufacturers.

• Delivered core full-stack functionality for a compliance-focused SaaS product under strict EU medical regulatory constraints.
• Built and refined features to improve audit readiness, traceability, and usability for regulated workflows.
• Enhanced production stability with better validation, runtime visibility, and issue detection.
• Worked across Java/Spring back-end and Angular front-end to ship and maintain end-to-end product features.
• Contributed to test coverage, observability, and platform reliability in production.

It was a Y Combinator-backed financial operations and bookkeeping product for crypto businesses, helping teams track on-chain activity and reconcile it into accounting workflows.

• Implemented end-to-end product features centered on on-chain data and accounting workflows.
• Led OpenSea integration for NFT assets, expanding support for blockchain-based financial activity.
• Integrated blockchain activity into internal flows for financial tracking and reconciliation.
• Built automated end-to-end test coverage to minimize regressions and boost release confidence.
• Improved reliability and delivery predictability via enhanced testing and back-end logic.
• Contributed production features in a fast-paced startup environment.

Products that help local governments and stakeholders understand property data, planning constraints, and development opportunities.

• Led a team of 5 engineers.
• Owned end-to-end delivery of two production applications.
• Coordinated technical execution across product requirements and engineering delivery.
• Contributed hands-on across back-end, front-end, and data-intensive workflows.
• Shaped architecture and implementation decisions for scalable geospatial products.
• Improved data reliability, delivery speed, and operational stability.
• Strengthened automation, maintainability, and engineering standards.

It's a telecom data platform project for Turkey's largest semi-government telecommunications company, focused on collecting, processing, generating signals, and reporting real-time and periodic telemetry from DSLAM and OLT devices across millions of customer lines.

• Developed microservices for collecting customer telemetry from DSLAM and OLT devices via SNMP.
• Built Kafka-based data flows for real-time and periodic collection at scale.
• Maintained Kubernetes-based infrastructure for collection and processing services.
• Supported high-volume ingestion pipelines serving millions of customer lines.
• Built internal JSF interfaces for operating and monitoring microservices.
• Investigated production issues to maintain stable live data collection.