Technical reports on the models and hardware we test, and blog posts about the systems we build.https://lokahq.github.io/tech-blog/en-ushttps://lokahq.github.io/tech-blog/running-large-language-models-fully-offline-on-mobile-with-react-native/https://lokahq.github.io/tech-blog/running-large-language-models-fully-offline-on-mobile-with-react-native/Paolo Pecis explores fully offline speech-to-text and LLM summarization on iOS and Android with React Native, whisper.rn, llama.rn, and the SoloAI app.Mon, 05 Oct 2026 00:00:00 GMTPaolo PecisReact NativeLLMMachine LearningMobile DevelopmentOffline AIhttps://lokahq.github.io/tech-blog/agentcore-identity/https://lokahq.github.io/tech-blog/agentcore-identity/Authentication and authorization for AI agents on AWS. Who is allowed to call your agent, and how your agent proves itself to everything it calls.Tue, 22 Sep 2026 00:00:00 GMTMatheus Dias2026-09-30T00:00:00.000ZAmazon Bedrock AgentCoreOAuthAI AgentsSecurityhttps://lokahq.github.io/tech-blog/esmc-trainium2/https://lokahq.github.io/tech-blog/esmc-trainium2/A native PyTorch benchmark showing EvolutionaryScale ESMC-300M running on AWS Trainium2: 490.8 variants per second on a single trn2.3xlarge — approximately 42.4 million protein variants scored per day at approximately $1.26 per million, 60% below H100 in this fixed-shape D2Deep benchmark.Tue, 18 Aug 2026 00:00:00 GMTJoão CorreiaTelmo FelgueiraTiago GonçalvesBojan JakimovskiJim BurtoftLouise Ping2026-09-30T00:00:00.000ZESM-CAWS TrainiumProtein Language ModelsBenchmarkinghttps://lokahq.github.io/tech-blog/a-prompt-pointed-the-other-way/https://lokahq.github.io/tech-blog/a-prompt-pointed-the-other-way/Everything is buildable now. Every team I meet lives in the same climate: more ideas than quarters to test them in, more tools than problems, roadmaps that read like wish lists because nothing on them is technically…Tue, 04 Aug 2026 00:00:00 GMTAna Marković2026-09-30T00:00:00.000ZProduct DesignDesign ThinkingAIWorkshop FacilitationProduct Developmenthttps://lokahq.github.io/tech-blog/benchmarking-grok-4-3/https://lokahq.github.io/tech-blog/benchmarking-grok-4-3/The main post covers what we found: Grok 4.3 lands in the same 96–97% accuracy band on GSM8K whether you reach it through Amazon Bedrock Mantle or the first-party xAI API, and low reasoning effort captures most of the…Fri, 10 Jul 2026 00:00:00 GMTNina CvetkovskaBojan Jakimovski2026-09-30T00:00:00.000ZGrokxAIAWSAmazon BedrockSpaceXhttps://lokahq.github.io/tech-blog/grok-bedrock-xai/https://lokahq.github.io/tech-blog/grok-bedrock-xai/A public-facing benchmark story on Grok 4.3 through Amazon Bedrock Mantle and the xAI API, focused on accuracy, latency, cost, reliability, and enterprise deployment tradeoffs.Fri, 10 Jul 2026 00:00:00 GMTNina CvetkovskaBojan Jakimovski2026-09-30T00:00:00.000ZGrokAmazon BedrockxAIBenchmarkinghttps://lokahq.github.io/tech-blog/contributing-to-open-source-bioinformatics-our-experience-at-the-nf-core-hackathon-from-medellin/https://lokahq.github.io/tech-blog/contributing-to-open-source-bioinformatics-our-experience-at-the-nf-core-hackathon-from-medellin/If you work anywhere near data-heavy science — genomics, metagenomics, proteomics — you’ve probably hit the same wall: building reliable, reproducible data pipelines is hard. Nextflow is an open-source workflow engine…Thu, 09 Jul 2026 00:00:00 GMTAndres FlorianDaniel Sabogal2026-09-30T00:00:00.000ZBioinformaticsNextflowMetagenomicsHackathonshttps://lokahq.github.io/tech-blog/from-proteins-to-pipelines-how-open-source-nextflow-tools-are-accelerating-drug-discovery/https://lokahq.github.io/tech-blog/from-proteins-to-pipelines-how-open-source-nextflow-tools-are-accelerating-drug-discovery/Modern computational drug discovery depends on complex workflows that connect AI models, structural biology tools, databases, and quality-control steps into reproducible pipelines. While many organizations have access…Wed, 08 Jul 2026 00:00:00 GMTJelena PejovicJorge Moura Sampaio2026-09-30T00:00:00.000ZNextflowMachine LearningDrug DiscoveryBioinformaticsOpen Sourcehttps://lokahq.github.io/tech-blog/computer-vision-at-scale-on-sagemaker-jobs-four-decisions/https://lokahq.github.io/tech-blog/computer-vision-at-scale-on-sagemaker-jobs-four-decisions/Training a vision model on millions of images is, more than anything, a data problem wearing an ML costume. The architecture can many times be the least interesting part. Everything around it is usually more fun: how we…Mon, 06 Jul 2026 00:00:00 GMTDidier2026-09-30T00:00:00.000ZComputer VisionData ScienceAIAmazon SageMakerBig Datahttps://lokahq.github.io/tech-blog/pushing-open-source-tts-models-to-their-limits-six-paradigms-14-models-one-production-reality/https://lokahq.github.io/tech-blog/pushing-open-source-tts-models-to-their-limits-six-paradigms-14-models-one-production-reality/In the digital audio subset of AI, the questions currently on everyone’s mind areThu, 02 Jul 2026 00:00:00 GMTErvin ShaqiriAlexandre Domingues2026-07-07T00:00:00.000ZOpen SourceTTSAudioDeploymenthttps://lokahq.github.io/tech-blog/aws-bedrock-guardrails-implementing-an-allowlist/https://lokahq.github.io/tech-blog/aws-bedrock-guardrails-implementing-an-allowlist/When building AI agents with AWS Bedrock, guardrails are your first line of defense for keeping conversations on topic. A common requirement is to restrict an agent to a pre-defined strict allowlist, blocking everything…Wed, 01 Jul 2026 00:00:00 GMTGuilherme Ribeiro2026-09-30T00:00:00.000ZAmazon Bedrock GuardrailsAI AgentsStrands AgentsAllowlistinghttps://lokahq.github.io/tech-blog/tts-emotion-benchmark/https://lokahq.github.io/tech-blog/tts-emotion-benchmark/Fourteen open-source text-to-speech models across six emotion-control paradigms, benchmarked against AWS Nova Sonic v2 on naturalness, expressiveness, latency, and licensing.Wed, 01 Jul 2026 00:00:00 GMTErvin ShaqiriAlexandre DominguesTTSNova SonicEmotion ControlBenchmarkinghttps://lokahq.github.io/tech-blog/running-six-open-source-cofolding-models-on-aws-lessons-learned-from-a-compute-perspective/https://lokahq.github.io/tech-blog/running-six-open-source-cofolding-models-on-aws-lessons-learned-from-a-compute-perspective/What to know before you scale cofolding jobs from five structures to ten thousand.Tue, 30 Jun 2026 00:00:00 GMTJulián F. FernándezAndres FlorianDaniel Sabogal2026-09-30T00:00:00.000ZAlphaFoldDrug DiscoveryAIBiologyChemistryhttps://lokahq.github.io/tech-blog/watching-a-python-to-rust-rewrite-was-painful-enough-to-build-this/https://lokahq.github.io/tech-blog/watching-a-python-to-rust-rewrite-was-painful-enough-to-build-this/How we built an agent that ports Python ML code to edge hardware for space.Fri, 26 Jun 2026 00:00:00 GMTJoão Afonso Pereira2026-09-30T00:00:00.000ZPythonRustEdge ComputingSpaceAI Agentshttps://lokahq.github.io/tech-blog/benchmarking-gpt-5-5-on-amazon-bedrock-vs-the-openai-api-the-engineering-walkthrough/https://lokahq.github.io/tech-blog/benchmarking-gpt-5-5-on-amazon-bedrock-vs-the-openai-api-the-engineering-walkthrough/The main post covers what we found: GPT-5.5 on Amazon Bedrock matches OpenAI API accuracy and comes out faster on every latency and throughput dimension we measured. This post covers how we ran it. If you want to…Mon, 15 Jun 2026 00:00:00 GMTPetar KalinovskiBojan Jakimovski2026-09-30T00:00:00.000ZAIAWSOpenAIMachine LearningLLMhttps://lokahq.github.io/tech-blog/openai-bedrock/https://lokahq.github.io/tech-blog/openai-bedrock/A comparison of GPT-5.5 through Amazon Bedrock and the OpenAI API across answer quality, latency, throughput, and reliability, followed by two production-shaped copilot builds.Mon, 15 Jun 2026 00:00:00 GMTPetar KalinovskiBojan Jakimovski2026-09-30T00:00:00.000ZOpenAIGPT-5.5Amazon BedrockEvaluationhttps://lokahq.github.io/tech-blog/carbon-trainium2/https://lokahq.github.io/tech-blog/carbon-trainium2/A practical benchmark showing Hugging Face Bio Carbon running on AWS Trainium2 with NxD Inference, covering what worked, what we measured, and why it matters for bio and HCLS teams moving open models into production.Thu, 21 May 2026 00:00:00 GMTBojan JakimovskiLoka Applied Research2026-09-30T00:00:00.000ZCarbonAWS TrainiumNxD InferenceGenomicshttps://lokahq.github.io/tech-blog/running-hugging-faces-carbon-on-aws-trainium2-with-nxd-inference/https://lokahq.github.io/tech-blog/running-hugging-faces-carbon-on-aws-trainium2-with-nxd-inference/Hugging Face Bio’s Carbon release is interesting for two reasons at once. First, it is a biology model. Second, it is not an infrastructure outlier. Many genomic models come with custom architectures, specialized…Thu, 21 May 2026 00:00:00 GMTBojan Jakimovski2026-06-14T00:00:00.000ZAWS TrainiumHugging FaceLLM InferenceAIAWShttps://lokahq.github.io/tech-blog/data-governance-in-practice/https://lokahq.github.io/tech-blog/data-governance-in-practice/Lineage for a CDC Pipeline on AWSMon, 04 May 2026 00:00:00 GMTCecilia Brusquetti2026-09-30T00:00:00.000ZData LineageAmazon MSKKafkaData Governancehttps://lokahq.github.io/tech-blog/beyond-static-rules-building-agentic-content-evaluation-systems-on-amazon-bedrock/https://lokahq.github.io/tech-blog/beyond-static-rules-building-agentic-content-evaluation-systems-on-amazon-bedrock/How to turn scattered guidelines into adaptive workflows that score, explain, and improve content at scaleWed, 22 Apr 2026 00:00:00 GMTCrhistian Cardona2026-09-30T00:00:00.000ZAWSAI AgentsGenerative AIAIAmazon Bedrockhttps://lokahq.github.io/tech-blog/ia-comedia-a-loka-made-lab-for-theatre-with-live-ai-actors/https://lokahq.github.io/tech-blog/ia-comedia-a-loka-made-lab-for-theatre-with-live-ai-actors/A Portuguese playwright wanted to push AI beyond the assistant role and onto the theatre stage. Here’s how LOKA built a live system that brings a cast of AI actors to perform alongside humans.Wed, 25 Mar 2026 00:00:00 GMTAndreia Pereira2026-09-30T00:00:00.000ZElevenLabsTheatreVoice Agentshttps://lokahq.github.io/tech-blog/the-agent-assembly-line/https://lokahq.github.io/tech-blog/the-agent-assembly-line/Productionizing Agentic Use Cases in Weeks, Not MonthsMon, 02 Mar 2026 00:00:00 GMTMario PetkoskiBojan JakimovskiZafir Stojanovski2026-09-30T00:00:00.000ZAWSLLMAI AgentsMachine LearningAIhttps://lokahq.github.io/tech-blog/deploying-trinity-mini-drugprot-think-on-amazon-sagemaker-ai/https://lokahq.github.io/tech-blog/deploying-trinity-mini-drugprot-think-on-amazon-sagemaker-ai/If you work in regulated domains (Healthcare, Life Sciences, Finance) you routinely hit constraints that break the default “just call a hosted API” approach:Mon, 23 Feb 2026 00:00:00 GMTBojan JakimovskiPetar Kalinovski2026-09-30T00:00:00.000ZArceeAmazon SageMakerLLMLife SciencesDrug Discoveryhttps://lokahq.github.io/tech-blog/trinity-drugprot/https://lokahq.github.io/tech-blog/trinity-drugprot/An ablation-driven study of GRPO-style reinforcement learning with LoRA on Trinity Mini for biomedical relation extraction, covering the training choices that materially changed performance.Mon, 23 Feb 2026 00:00:00 GMTBojan JakimovskiPetar Kalinovski2026-09-30T00:00:00.000ZRLVRLoRAMixture of ExpertsDrugProthttps://lokahq.github.io/tech-blog/clean-architecture-with-strategy-pattern-in-kotlin/https://lokahq.github.io/tech-blog/clean-architecture-with-strategy-pattern-in-kotlin/When building stable and maintainable software, one of the biggest challenges developers face is that modern applications often need to support multiple versions of a feature—different algorithms or formats that can…Tue, 27 Jan 2026 00:00:00 GMTStefanija ZdraveskaAndroidStrategy Patternhttps://lokahq.github.io/tech-blog/building-rag-systems-on-aws-lessons-from-serverless-and-ec2-benchmarks/https://lokahq.github.io/tech-blog/building-rag-systems-on-aws-lessons-from-serverless-and-ec2-benchmarks/A practical benchmark of RAG ingestion and search performance across AWS Lambda and EC2Tue, 20 Jan 2026 00:00:00 GMTRomulo PagnozziCrhistian Cardona2026-09-30T00:00:00.000ZAWSRAGVector DatabaseBenchmarkingServerlesshttps://lokahq.github.io/tech-blog/rethinking-bioinformatics/https://lokahq.github.io/tech-blog/rethinking-bioinformatics/Solving the Challenges of Traditional PipelinesMon, 05 Jan 2026 00:00:00 GMTSara Oquendo2026-09-30T00:00:00.000ZBioinformaticsCloud ComputingData EngineeringScientific WorkflowsConnectedlabhttps://lokahq.github.io/tech-blog/part-2-agentic-patterns-101-with-loka-strands-agents/https://lokahq.github.io/tech-blog/part-2-agentic-patterns-101-with-loka-strands-agents/After exploring the foundations of Agentic Patterns in Part 1, where we looked at more structured workflows like Sequential and Parallel workflows, LLM Routing, and Reflection, this time we are going to push further…Tue, 02 Sep 2025 00:00:00 GMTNina CvetkovskaPetar KalinovskiBojan Jakimovski2026-09-30T00:00:00.000ZStrands AgentsAgentic AIGenerative AIAmazon BedrockAI Agentshttps://lokahq.github.io/tech-blog/part-1-agentic-patterns-101-with-loka-strands-agents/https://lokahq.github.io/tech-blog/part-1-agentic-patterns-101-with-loka-strands-agents/These days, Agents are everywhere in daily conversation and it seems everyone wants one. We read or hear about AI Agents, Agentic AI, Agentic Architectures, and Agentic Patterns almost every day. But what do these terms…Wed, 27 Aug 2025 00:00:00 GMTNina CvetkovskaPetar KalinovskiBojan Jakimovski2026-09-30T00:00:00.000ZStrands AgentsAgentic AIGenerative AIAI AgentsAmazon Bedrockhttps://lokahq.github.io/tech-blog/flattening-the-curve-with-nextflow-building-a-scalable-and-reproducible-bioinformatic-workflow-for/https://lokahq.github.io/tech-blog/flattening-the-curve-with-nextflow-building-a-scalable-and-reproducible-bioinformatic-workflow-for/Loka’s participation in the recent nf-core hackathon in Medellín, Colombia, culminated in the successful development of MITNANEX, a pipeline designed to extract and assemble mitochondrial genomes, identify and annotate…Sat, 28 Jun 2025 00:00:00 GMTJelena PejovicJosé RibónAndrés SacreFederico RuedaNicolás FrancoJavier DominguezJuliana Silva2026-09-30T00:00:00.000ZBioinformaticsData EngineeringNextflowOxford NanoporeGenomicshttps://lokahq.github.io/tech-blog/how-to-transition-from-xml-android-view-to-jetpack-compose-view/https://lokahq.github.io/tech-blog/how-to-transition-from-xml-android-view-to-jetpack-compose-view/Moving from XML to Jetpack Compose helps developers write cleaner, faster and more modern UI code — essential for staying up to date with Android’s best practices.Mon, 23 Jun 2025 00:00:00 GMTGabriel Menezes da Silva2026-09-30T00:00:00.000ZAndroidJetpack ComposeAndroid Viewshttps://lokahq.github.io/tech-blog/aws-s3-browser-will-save-you-time/https://lokahq.github.io/tech-blog/aws-s3-browser-will-save-you-time/Simple interface for data stored in S3Tue, 17 Jun 2025 00:00:00 GMTDragan Savevski2026-09-30T00:00:00.000ZAWSS3Amplifyhttps://lokahq.github.io/tech-blog/part-4-benchmarking-deepseek-cost-performance-and-business-use-cases-on-aws/https://lokahq.github.io/tech-blog/part-4-benchmarking-deepseek-cost-performance-and-business-use-cases-on-aws/A DeepSeek benchmark on AWS that compares cost, performance, and deployment choices for business workloads using open-weight models.Thu, 10 Apr 2025 00:00:00 GMTCrhistian CardonaBojan Jakimovski2026-09-30T00:00:00.000ZAIGenerative AIDeepSeekOpen SourceAWShttps://lokahq.github.io/tech-blog/part-2-deploying-distiled-deepseek-r1-models-on-amazon-sagemaker-ai/https://lokahq.github.io/tech-blog/part-2-deploying-distiled-deepseek-r1-models-on-amazon-sagemaker-ai/A deployment walkthrough for serving distilled DeepSeek-R1 models with Amazon SageMaker AI, including the AWS setup for open-weight reasoning models.Thu, 20 Feb 2025 00:00:00 GMTCrhistian CardonaBojan Jakimovski2026-09-30T00:00:00.000ZAWSGenerative AIAIDeepSeekhttps://lokahq.github.io/tech-blog/part-3-deploying-deepseek-r1-models-on-aws-designed-silicon-instances/https://lokahq.github.io/tech-blog/part-3-deploying-deepseek-r1-models-on-aws-designed-silicon-instances/Engineering notes on deploying DeepSeek-R1 models to AWS-designed silicon instances and serving open-weight reasoning models on AWS.Thu, 20 Feb 2025 00:00:00 GMTCrhistian CardonaBojan Jakimovski2026-09-30T00:00:00.000ZAWSGenerative AIAIDeepSeekhttps://lokahq.github.io/tech-blog/harnessing-open-source-ai-on-aws/https://lokahq.github.io/tech-blog/harnessing-open-source-ai-on-aws/Deploy distilled DeepSeek-R1 open-weight models through Amazon Bedrock, with a hands-on AWS setup from the Loka engineering team.Mon, 17 Feb 2025 00:00:00 GMTCrhistian CardonaBojan Jakimovski2026-09-30T00:00:00.000ZAWSGenerative AIAIAmazon BedrockDeepSeekhttps://lokahq.github.io/tech-blog/dive-into-deepseek/https://lokahq.github.io/tech-blog/dive-into-deepseek/How the open-recipe LLM is transforming GenAIWed, 12 Feb 2025 00:00:00 GMTCrhistian CardonaBojan Jakimovski2026-09-30T00:00:00.000ZDeepSeekLLMGenerative AIAWShttps://lokahq.github.io/tech-blog/how-http-3-0-and-quic-solve-head-of-line-blocking/https://lokahq.github.io/tech-blog/how-http-3-0-and-quic-solve-head-of-line-blocking/You probably have heard the history of how HTTP was created and you might even know about the new HTTP/3 version. In this article, I want to focus on some of the lesser-known details of HTTP and why we need a new…Mon, 20 Jan 2025 00:00:00 GMTDaniel Pereira2026-09-30T00:00:00.000ZNetworkingSoftware Engineering