Skip to content
View KooshaPari's full-sized avatar

Highlights

  • Pro

Block or report KooshaPari

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
KooshaPari/README.md

👋

Software engineer working across systems software, distributed and agent infrastructure, ML inference/evaluation, developer tooling, and cloud platforms. I also have a background in technical product/program leadership and physical-product commercialization.

AI slop inside GitHub Downloads (all assets, all releases)

Selected Engineering Work

Rust/Tokio OS-adjacent runtime for high-concurrency coding-agent workloads, including process supervision, resource-pressure awareness, duplicate-command coalescing, thermal/contention gating, optional FUSE interception, health checks, and Prometheus telemetry.

Rust/Go AI dispatch and routing substrate with streaming provider routing, rate limiting, retries, fallback chains, circuit breakers, per-session budgets, audit logging, hot-reloaded configuration, and HTTP/CLI/MCP interfaces.

OMLX-derived local-inference research stack extending MLX with multi-backend routing, Rust performance cores, speculative decoding, quantization, concurrent execution, model evaluation, and Apple-Silicon kernel experiments.

Declarative GitHub-to-cloud deployment platform with authenticated Go APIs, AWS provisioning, service registration, and deployment observability.

Rust traceability and audit infrastructure connecting software artifacts, evidence, agent activity, and delivery state across server, CLI, edge, and MCP interfaces.

Extended open-source multi-provider AI proxy work covering provider-aware routing, authentication, quotas, operational controls, diagnostics, and integration with broader agent infrastructure.

Additional Work

External Open Source Contributions

OmniRoute — External Contributor

Rank #5 in published upstream contributor census · 101 merged PRs

Contributed across routing intelligence, provider integrations, reliability hardening, API/protocol compatibility, and operational tooling.

Highlights:

  • Bifrost auto-fallback cooldown and latency/speed-optimized routing
  • OpenAPI 3.0 + Redoc API documentation
  • Provider-manifest infrastructure and integrations (Factory.ai, MiniMax M3, xAI, Bailian, OpenAI-compatible MCP Responses)
  • Reliability hardening across cooldowns, fallback caches, and stream lifetime
  • Contributions named in 21 upstream releases; personally acknowledged by the maintainer

View merged contributions → upstream PR search

Earlier Work

Frostify — substantially customized Spicetify theme released in 2020; packaged GitHub release accumulated 3,350+ release-asset downloads.

Pinned Loading

  1. Tracera Tracera Public

    Rust traceability infrastructure: audit trails and provenance for agent workflows

    TypeScript

  2. BytePort BytePort Public

    Declarative Go/AWS deployment platform: GitHub repo to running service

    Svelte

  3. zz-aa-dep-CliproxyApiPlusPlus zz-aa-dep-CliproxyApiPlusPlus Public

    Forked from router-for-me/CLIProxyAPI

    Fork dep: multi-provider AI proxy with routing, auth, quotas for OmniRoute

    HTML 36

  4. OmniRoute OmniRoute Public

    Forked from diegosouzapw/OmniRoute

    Multi-provider LLM gateway: OpenAI-compatible endpoint with smart routing across 300+ models

    TypeScript 14

  5. ShareCLI ShareCLI Public

    Rust process/resource runtime: high-concurrency agent workload scheduler

    Rust 1

  6. gwennlbh/Frostify gwennlbh/Frostify Public

    A theme for spicetify focused on being transparent

    CSS 7