Karlis Vigulis

• In 2 years working for the business, I've built a feature rich vacation rental platform with excellent UI & UX (https://www.abodevillas.com). A super efficient world-class serverless architecture with Golang backend, raw SQL (full control, and no more unpredictable query overhead from ORMs) with sub 100 milliseconds response times.
• Carefully architected the whole system to ensure full control, no vendor lock-in, top notch performance, very cost-efficient usage based infrastructure, with heavy focus on maintaining pleasant DX. For this use case, I have become a big fan of end-to-end type safety and DX offered by tRPC + postgreJS + Atlas (for DB migrations) setup deployed on AWS lambda via SST IaC framework.
• Designed secure AWS infrastructure adhering to GDPR and PoLP principles: custom VPC segmentation, restricted inter-service security groups, least-privilege IAM policies, and comprehensive audit logging.
• Stack: NextJS; SST; AWS Lambda; Golang BE (for website for maximum performance), TS+tRPC+zod (for internal admin system for best DX and great form validation); Postgres; SES; AWS Cognito; Atlas for DB schema management.

Notable features:

• Google Vacations Rentals integration which saves the company about £8-15/month per live property, compared to using a 3rd party PMS, as well as 10+ API integrations with other partners (villa suppliers).
• Very modern CMS & CRM admin system with fine-tuned AI assistants for multiple interfaces, which can take raw text data and reliably serialize it into structured data fed to the admin forms, thus resulting in 5-10x faster data entry and validation turnover times over fully manual legacy UI.
• Internal tooling to do image search based on cosine similarity over CNN image embeddings done very efficiently using good old Postgres via pgvector extension.

• Self-motivated, with a strong sense of personal responsibility.
• Excellent communication skills, both verbal and written.
• Proven ability to learn quickly and adapt to new situations.
• Skilled at working independently and collaboratively in a team environment.

• Worked across the tech stack, primarily with Django, ReactJS and Golang.
• Created numerous new features for the CRM and know my way around Django admin classes and customisation.
• Created custom reporting dashboards, a WYSWIG-like pitch page app that can be tailored to target specific customers.
• Rewrote pages using ReactJS and NextJS to move them out of a legacy monolith into containerised micro-services and greatly improved the SEO performance metrics as a result.
• Played an integral part in gradual migration of legacy business logic (Django) and front-end (ReactJS) app over to a micro-service based architecture (based on Golang and ReactJS + NextJS).
• Gained experience with mono-repo micro-service projects, deployed using K8s + Helm + Treafik ingress, and notably on how to handle routing and data storage for multiple/parallel staging deployments.
• Part of a work environment adhering to Agile Scrum methodology.

• Worked on a variety of projects involving large scale image retrieval, semantic & instance segmentation and SLAM.
• Responsible for creating an image search / GeoAnalytics web platform (Django as BE and plain JavaScript at FE)
• ML frameworks - primarily PyTorch and scikit-learn.

• Team member of an early start-up, building a Combined Heat and Power gasification plant.
• Oversaw the electronics/automation projects, microcontroller programming and a real-time monitoring software development (C and C++).