Skip to content
View sharmis-tp's full-sized avatar

Block or report sharmis-tp

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. 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
sharmis-tp/README.md

Senior Rust Engineer · AI/ML Infrastructure

The systems layer under machine learning: training libraries, checkpoints, agent orchestration, LLM observability and developer tooling.

Rust Go Python TypeScript OpenTelemetry

followers email open to remote work


whoami

I've been building software since 2018, most of it in Rust: async services, FFI, memory allocators and protocol code that has to stay correct under load.

Today I work where that background matters most in AI: the infrastructure that models and agents run on. That means numerical correctness in training code, checkpoints that survive a crash, evaluation scores you can trust, and traces that tell you why an agent did what it did.

Models are only as reliable as the systems that train, evaluate and serve them.

What I work on

Area What I focus on
Training library internals Optimizer math (RMSprop variance, epsilon placement), weight initialization (Kaiming/Xavier), RNN input validation, numerically stable edge cases
Checkpointing & data integrity Atomic checkpoint writes, batched tensor serialization, verified dataset downloads
Agent orchestration & safety Lock lifecycle in task orchestrators, eval-dimension scoring, per-agent event types, scope enforcement
LLM observability OpenTelemetry spans for agent and tool events, error capture in LangChain tracing, token-usage parsing, AI memory and knowledge planes
Grafana & monitoring ECharts plugin hardening, template variable regex, Pyroscope datasource fixes
Forecasting & anomaly detection Exogenous-feature support in time-series foundation-model pipelines
Systems foundations Tokio, cancellation safety, backpressure, zero-copy I/O, FFI and unsafe audits

Open-source contributions

All links go to my own pull requests. 55+ merged across 30+ repositories.

AI/ML & agent infrastructure

Project Contribution Status
ClickHouse/loom: AI agent memory & observability on ClickHouse Rewrite doc page opening sentences for search and LLM discovery merged
FastNN: neural-network library in Rust + CUDA Reject zero-length sequences in LSTM/GRU · Assert fan_in > 0 in Kaiming/Xavier init merged
Atomic checkpoint writes · Verify MNIST download integrity merged
jamjet: safety layer for AI agents Fix DimensionScores.composite() weight normalization · Add OTel span attributes merged
forge-orchestrator: agent task orchestration Release orphaned locks for completed tasks · Fix companion plugin on re-run merged
PrismorSec/prismor: security agent platform Detect --mount bind in host policy · Declare per-agent event_types · Synthesized scope asks instead of blocking merged
durare: provider framework Add StepOutcome::from_columns · Export types named by StateProvider methods merged
Sentient: AI agent desktop Keep names like Claude Code in generated setting labels merged
automic-vault: agent scripting Add agent skills for Blessed Scripts merged

Observability & Grafana

Project Contribution Status
grafana/dataViz: Grafana ECharts plugin monorepo Guard non-string config.src in getLayerSrc · Replace untracked plan file references in review
grafana-pyroscope-datasource Fix multi-value template variable regex in review
getsentry/sentry-go: Sentry SDK for Go Add WithMechanism capture option in review
lemma: LLM tracing Record handled LangChain tool errors as span errors in review
oxtrace: LLM usage tracing Parse usage from gzip-encoded responses in review
foundationforecast: time-series foundation models Forward futr_exog_list in detect_anomalies in review

Systems, networking & blockchain

Project Contribution Status
libp2p/py-libp2p: P2P networking in Python Use /sni/ name as TLS server name when dialing merged
libp2p/js-libp2p: P2P networking in JS Wrap ErrorEvent in ConnectionFailedError on failed dial · Validate config options at startup in review
polkadot-api/papi-console: Polkadot console Re-run runtime calls from workspace · Keep fork lanes stable merged
UOR-Foundation/uor-addr: content addressing in Rust Reject JSON objects with duplicate member names · Sort object members by UTF-16 code units merged
UOR-Foundation/kappa-registry: OCI registry in Rust Carry page size in tags/list Link header · Add dual MIT/Apache license in review
project-llzk/circom: ZK circuit compiler Add location and expression to assertion messages merged
perun-network/perun-ckb-backend: payment channels Make subscription polling interval configurable merged
ScopeLift/umbra-protocol: privacy protocol Remove unimplemented relayWithdrawOnBehalf() stub merged
MorpheusAIs/Morpheus-Lumerin-Node: decentralized AI Re-check sessions on chain before refusing in review
stellar-scaffold/cli: Stellar dev CLI Build clients without a Cargo workspace in review

Developer tools & language servers

Project Contribution Status
eza-community/eza: modern ls replacement Add EZA_SIZE_STYLE env var in review
beancount-language-server: LSP for Beancount Advertise positionEncoding capability (LSP 3.17) · Fix exit code on unclean shutdown merged
tree-sitter-beancount: grammar Add injections.scm for BQL query highlighting in review
metal2vk: Metal-to-Vulkan translation Key pipeline variants by library/name/constants · Reject inconsistent push/descriptor metadata merged
smolvm: lightweight VM manager Let a disk volume set its directory's owner and mode in review
mimas: GML parser in Rust Accept loop, while, for, if and match as binary operands merged

Own repositories: smalloc (memory allocator for Solana programs), Dbgeng-sys (Rust FFI bindings for the Windows debugger engine), Solana-Credit-Score, AsyncCSVX.

How I work

01. Measure, then optimize. Throughput, tail latency and allocations, not guesses.

02. Make failure boring. Writes are atomic, inputs are validated at the boundary, and errors say what to do next.

03. Trust numbers only when they're tested. An eval score or a loss curve is only as good as the code that computes it.

04. Small, reviewable changes. Every PR above is scoped to one problem and easy to review.


Open to remote roles and contracts with AI/ML startups: ML infrastructure, inference, evaluation and Rust performance work.

Pinned Loading

  1. awesome-rust awesome-rust Public

    Forked from rust-unofficial/awesome-rust

    A curated list of Rust code and resources.

    Rust 2 1

  2. Solana-Credit-Score Solana-Credit-Score Public

    Solana Credit Score shows which validators give the best staking rewards.

    Rust 1

  3. omnibus omnibus Public

    This repo contains code for managing and testing token-related programs

    Rust

  4. cross cross Public

    Forked from cross-rs/cross

    “Zero setup” cross compilation and “cross testing” of Rust crates

    Rust

  5. slint slint Public

    Forked from slint-ui/slint

    Slint is an open-source declarative GUI toolkit to build native user interfaces for Rust, C++, JavaScript, or Python apps.

    Rust

  6. smalloc smalloc Public

    A small memory allocator for Solana programs

    Rust