Looks like you have JavaScript disabled. For the full Twine experience, you will need to re-enable it.

I’m Daniel Michalski, a Senior Software Engineer based in Warsaw, Poland, focused on AI/ML platforms and real-time inference. I design fault-tolerant Go services, event-driven data pipelines, and scalable model serving with strong SRE, security, and observability practices. My work spans experimentation through registry, online/offline features, shadow/canary deployments, and monitoring for drift. I have a track record of cutting latency, improving fraud precision, and turning ML prototypes into production systems. I enjoy turning ideas into reliable software, driving cross-team collaboration, and mentoring engineers to raise reliability and performance across the stack.…I’m Daniel Michalski, a Senior Software Engineer based in Warsaw, Poland, focused on AI/ML platforms and real-time inference. I design fault-tolerant Go services, event-driven data pipelines, and scalable model serving with strong SRE, security, and observability practices. My work spans experimentation through registry, online/offline features, shadow/canary deployments, and monitoring for drift. I have a track record of cutting latency, improving fraud precision, and turning ML prototypes into production systems. I enjoy turning ideas into reliable software, driving cross-team collaboration, and mentoring engineers to raise reliability and performance across the stack.

Daniel Michalski

Full Stack Developer, Web Developer, Programmer, +2





I’m Daniel Michalski, a Senior Software Engineer based in Warsaw, Poland, focused on AI/ML platforms and real-time inference. I design fault-tolerant Go services, event-driven data pipelines, and scalable model serving with strong SRE, security, and observability practices. My work spans experimentation through registry, online/offline features, shadow/canary deployments, and monitoring for drift. I have a track record of cutting latency, improving fraud precision, and turning ML prototypes into production systems. I enjoy turning ideas into reliable software, driving cross-team collaboration, and mentoring engineers to raise reliability and performance across the stack.…I’m Daniel Michalski, a Senior Software Engineer based in Warsaw, Poland, focused on AI/ML platforms and real-time inference. I design fault-tolerant Go services, event-driven data pipelines, and scalable model serving with strong SRE, security, and observability practices. My work spans experimentation through registry, online/offline features, shadow/canary deployments, and monitoring for drift. I have a track record of cutting latency, improving fraud precision, and turning ML prototypes into production systems. I enjoy turning ideas into reliable software, driving cross-team collaboration, and mentoring engineers to raise reliability and performance across the stack.

Available to hire

I’m Daniel Michalski, a Senior Software Engineer based in Warsaw, Poland, focused on AI/ML platforms and real-time inference. I design fault-tolerant Go services, event-driven data pipelines, and scalable model serving with strong SRE, security, and observability practices. My work spans experimentation through registry, online/offline features, shadow/canary deployments, and monitoring for drift. I have a track record of cutting latency, improving fraud precision, and turning ML prototypes into production systems. I enjoy turning ideas into reliable software, driving cross-team collaboration, and mentoring engineers to raise reliability and performance across the stack.

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Intermediate

Intermediate

Intermediate

Intermediate

Language

English

Fluent

Work Experience

Senior Software Engineer at Coinbase

August 1, 2024 - November 13, 2025

Led a Go gateway around vLLM/Triton with online features (Feast + Redis) and sticky Kafka ingestion; achieved p99 latency reduction (-38%) and fraud precision improvement (+7.9%) behind a shadow/canary rollout. Built a model-aware risk service (gRPC gateway, gobreaker, token bucket) reducing chargeback losses by 11% QoQ with deterministic audit logs. Rolled out RAG for KYC/explanations and partner Q&A (Go control plane, Python inference, pgvector/OpenSearch) improving handoffs by 32% and answer accuracy by 18 percentage points. Governed protobuf schemas (buf) and fenced producers, reducing cross-AZ chatter by 35% and consumer lag by 80%. Instrumented observability with OpenTelemetry + Evidently + Datadog alerts-as-code, enabling automated rollback on AUC/FPR regression; MTTR reduced by 45% on trade-path incidents. Validated failure budgets via game-days/chaos to ensure backpressure and graceful degradation under bursty load. Hardened partner APIs for Advanced/Prime/NTX with per-tenant

Staff Software Engineer at GlobalLogic

June 1, 2024 - June 1, 2024

Launched an inference gateway (Go over Python: XGBoost/LightGBM→PyTorch) with real-time features via Kafka + Feast; achieved CTR +6.1% and conversions +4–6%. Upgraded search to hybrid (pgvector/OpenSearch + BM25) with NLP intents/NER and re-rankers; +14% success, 0-result −27%. Deployed an LLM chatbot with tool-calling/guardrails; ticket deflection ~28%, hallucinations <2% on evals. Standardized a Go service template (zap, Viper, health/ready) and CI/CD with blue/green + canary + auto-rollback; time-to-market −60%, p99 −35%, errors −40%, MTTR −30%. Coordinated CDC/ETL (Debezium, Airflow) and batch/stream (Spark/Flink, dbt, Delta) with OTEL traces FE ↔ BE ↔ ETL; freshness <15 min, downstream failures −30%. Fortified auth (OAuth/OIDC, mTLS) and supply-chain security (Trivy, SAST/DAST) with zero high-severity findings. Drafted RFCs for API versioning, schema evolution, and Kafka partitioning/compaction; pass rate >95% and reduced review defects ~30%. Mentored 5–7 eng

Software Engineer at Avra Software

February 1, 2020 - February 1, 2020

Built and maintained 25+ backend/integration services (Go) with event-driven workflows (Kafka/RabbitMQ), gRPC/REST, PostgreSQL/Redis; achieved 99.95% reliability with circuit breakers, idempotency, and blue-green deploys; cut p95 latency by ~45% via profiling, async I/O, and caching. Shipped full-stack features (React/Next.js, Angular, Vue) and a shared TS design system + API SDKs, reducing duplicate work by ~40% and trimming client round-trips by ~60%. Migrated ML from batch to near real-time with Python ETL + MLflow and Kafka features; delivered OCR/NLP POCs reducing manual reviews by ~40%. Ran platform & reliability for 3 Kubernetes clusters and 100+ CI/CD pipelines (Argo CD); improved MTTR and reduced costs via autoscaling and optimized storage. Strengthened observability (Prometheus/Grafana/OpenTelemetry) and incident playbooks, halving on-call time.

Intern Software Engineer at Google

May 1, 2014 - May 1, 2014

Shipped caching and gRPC batching on a Kubernetes service; learned Kubernetes, tracing, and clean PR practices; owned feature end-to-end and contributed to latency improvements.

Intern Software Engineer at Amazon

December 1, 2013 - December 1, 2013

Built an AWS Lambda/SQS/DynamoDB event pipeline with idempotency and dashboards; gained experience in event-driven design and AWS operations; collaborated across teams to reduce errors.