I like machines close to the metal and results you can measure. I run tiyuvta, a research lab that deploys open-source frontier models on hardware you can afford, and the serving engine that makes that possible is the core of the work. I also build in-memory data systems at AWS ElastiCache, and maintain open-source tools people actually run.

Research notes and working papers live at avifenesh.ai. Issues, questions, and counterexamples are always welcome.

tiyuvta is my research lab, and it sells one thing: self-deploying open-source frontier models on hardware you can afford, cheaper and faster than the stock recipe. The engine is the core of it: we rework the serving engine and the model's architecture so a frontier model fits the cards you can buy and runs fast there, measured, not promised. Around that: choosing the right models for your load (chat, embedding, rerank, specialists: what fits and what doesn't), fine-tuning on your workload, cloud setup when you'd rather not run hardware, days on site when it helps. The researcher who builds the deployment is the one who keeps supporting it. The lab's research below (engine, quantization, pruning, speculative decoding, Hebrew models) is what makes those deployments cheaper. We run one deployment ourselves as the proof, DeepSeek-V4.1-Flash on our own engine: put a few dollars in and test your workload against it before we talk. Examples: tiyuvta-use-cases · llm-tiyuvta.

• memra: tiyuvta's from-scratch Rust + CUDA research engine for RTX PRO 6000 Blackwell and RTX 5090, the instrument behind the studies below and the byte-exact oracle for the serving stack we deploy. Safetensors is the tuned path, GGUF stays supported, and a mechanism that wins on one card and loses on the other becomes a per-device default rather than a compromise. On crates.io with prebuilt binaries.
• hqmtp: MTP draft-head lab, concluded. Function cuts (pruning, low-rank, distillation) pay a 10–19-point off-distribution tax that fidelity cuts don't; the zero-training trimmed-vocabulary recipe won at 1.8–2.7× end to end. The negative results stay in the ledger.
• recipe-lab: layer-loop weight sharing + ε=λ/(N√L) residual scaling, combined for the first time and tested from zero in 11 pre-registered rounds. In the data-constrained regime the looped model beat FLOPs-matched vanilla in all three mixer families (attention, pure SSM, and hybrid; seven paired runs, zero sign flips), with 26–34% fewer parameters. Rule isolated: loop the state-mixer, never the retriever.
• mem-retrofit: grafted a product-key memory layer onto a stock dense 4B and ran it against LoRA over sequential updates. The retrofit is free at lr/10; the published forgetting advantage failed 12/12 confidence intervals.
• More studies with receipts: fixed-compute-frontier (a preregistered kill-gate ledger, ~84 theory lanes) · gemma-expert-atlas (26B MoE expert surgery, 3,840 experts traced) · block-routed-swiglu (near-free kernel, refuted capability) · moe-lab · assumption-excavator.
• Working papers: small-vocabulary MTP heads · prune, heal, quantize. Methods, failed arms, and evidence in the open.
• In review upstream: NVFP4 quantization at llama.cpp, hybrid-KV loads at vLLM and LMCache.

• Valkey and its ecosystem. I maintain Valkey GLIDE, the official multi-language client (Rust core, Java/JNI, Node/N-API), and valkey-skills, the official AI skills for the ecosystem, which I started. I contribute to Valkey itself, and a good part of my open-source time goes to the people around it: the clients team, talks, and helping the people who run it.
• agent-sh, my org: an ecosystem of tools for agent-assisted development, working across Claude Code, Codex, OpenCode, Cursor, and Kiro.
• glide-mq: Node.js queue on Valkey Streams with a Rust N-API core, plus adapters for Hono, Fastify, Hapi, NestJS and a dashboard.
• Also around: RustOwl (runtime, memory, and CI work) and ChatGPT Desktop for Linux (Wayland input, accessibility, packaging).

• Valkey: core contributor; sync-from-replica replication in review. On the side: CRIU copy-on-write live-migration research, under 50 ms of freeze while migrating a 200 GB loaded instance.
• FlowFabric: durable-execution engine in Rust for Valkey, Postgres, and SQLite: lease-safe workers, waitpoints, budgets.
• layout-audit: DWARF memory-layout analysis: padding, layout diffs, size budgets for C/C++/Rust/Go.
• scrump: format-aware secret scrubber for binary capture artifacts: perf.data, core dumps, nsys traces, JFR.
• ocaml-valkey: OCaml 5 + Eio Valkey client, published on opam.

• agnix: linter and language server for AI agent configs: 444 rules with autofixes, a GitHub Action, an MCP server, and editor plugins.
• computer-use-linux / agent-workspace-linux: Linux desktop control over MCP, and isolated agent-owned desktops so an agent never has to touch your real machine.
• eigen: local agent daemon in Go: resumable sessions, multi-provider routing, memory, observability.

• Talks: Inside Valkey GLIDE on the AWS Developers Podcast · Glide into resiliency on Let's Talk About Data
• Writing: avifenesh.ai/writing · answering on Stack Overflow
• 📫 aviarchi1994@gmail.com · LinkedIn · X

If something here saved you time, sponsoring helps me keep doing it.