Resume

Sam Parhimchik

Portrait of Sam Parhimchik

Hands on Engineering Manager and technical leader with senior engineering depth in Solana, fintech, developer infrastructure, and early stage product development.

At Orca I was promoted from Senior Software Engineer to Engineering Manager, leading three engineers and a product designer while remaining involved in architecture, implementation, analytics, and production systems.

Senior Software Engineer → Engineering Manager @ Orca · Aug 2024–Jul 2026

US citizen · Remote preferred · Open to relocation

Currently based in Miami

Review leadership experienceView selected case studiesDownload résumé

Experience

Engineering management with technical depth

Promoted from Senior Software Engineer to Engineering Manager at Orca. Led a team of three engineers and a product designer while remaining involved in architecture, implementation, and public developer infrastructure.

Orca · Engineering Manager

Senior Software EngineerEngineering Manager

Senior Software Engineer Aug 2024–Sep 2025Engineering Manager Sep 2025–Jul 2026

Managed three engineers · partnered with one product designer · Los Angeles | New York | Miami

Aug 2024 — Jul 2026

Solana liquidity layer. Promoted from Senior Software Engineer to Engineering Manager. As a senior engineer I led Wavebreak and shipped public transaction tooling; as Engineering Manager I led three engineers and a product designer across Vaults, analytics, and partner integrations while remaining involved in architecture and implementation.

  • Promoted from Senior Software Engineer (Aug 2024–Sep 2025) to Engineering Manager (Sep 2025–Jul 2026). Managed three engineers and partnered with one product designer.
  • As Senior Software Engineer, led Wavebreak from research through production — coordinating technical scope, product dependencies, delivery risk, and release sequencing immediately before the promotion.
  • Authored @orca-so/tx-sender (TypeScript) and contributed to orca_tx_sender (Rust): priority fees, Jito tips, compute estimation, retries, and confirmation. Reached more than 44K weekly npm downloads at peak.
  • Led the Kamino Vaults partner integration — automated yield strategies live at orca.so/vaults — coordinating Orca-owned product surfaces with external vault infrastructure.
  • Built Mixpanel instrumentation and funnels for concentrated liquidity deposits, identified abandonment in the deposit path, and used that analysis to change what the team prioritized next.
Management leverage
  • Managed three engineers across Vaults and platform work, setting ownership boundaries, sequencing concurrent releases, and staying involved on high-risk technical paths.
  • Owned planning and prioritization for the team, balancing roadmap commitments, partner dependencies, unplanned production work, and technical risk with one product designer.
  • Ran hiring and interviewing for engineering roles on the team.
  • Held recurring one-on-ones and performance feedback with three engineers, pairing coaching with design review on the work they owned.
  • Coordinated delivery risk during the Kamino Vaults partner integration, tracking external dependencies and release sequencing through production.
  • Introduced LLM-assisted TDD and PR review on the team: structured prompts, CI-run generated tests, and human merge approval. Review comments were treated as input, not as a substitute for engineering judgment.
Team and product ownership

As Engineering Manager I was responsible for getting the team's products delivered: planning, prioritization, technical judgment, partner coordination, and release sequencing. Individual contributors owned feature implementation. The product designer owned interaction and visual design. I stayed close enough to the systems to make technical calls and contribute directly when the risk warranted it.

Wavebreak was the main product I led as Senior Software Engineer, immediately before the promotion — Orca's meme token trading platform — from market research and user interviews through production, including on-demand infrastructure for vanity address generation.

Orca Vaults was a partner-backed product after that: Kamino operated the vault infrastructure; the team shipped Orca's deposit, withdraw, portfolio, and launch surfaces so allocators could use those strategies without leaving the platform.

Technical contribution

As Senior Software Engineer I designed and shipped public TypeScript transaction infrastructure and contributed to the companion Rust crate, before the promotion.

  • Authored @orca-so/tx-sender (TypeScript): transaction construction, priority fees, Jito tips, compute estimation, signer handling, confirmation, retries, and failure handling
  • Contributed companion Rust crate orca_tx_sender in the same public monorepo
  • Integrated the packages into the orca-so/whirlpools SDK ecosystem through a public pull request with iterative review (approximately 36 merged commits)
Analytics and product decisions

I instrumented concentrated liquidity deposit journeys in Mixpanel and built funnels and dashboards so the team could see where users abandoned the flow. The analysis changed what we built next: retention work on the deposit path took priority over adding more surface area.

OpenHaus · Co-founder · Technical Founder · Part time

Jan 2022 — 2024 · Vancouver, BC

Commission-free real estate marketplace I co-founded. I designed and implemented the product and infrastructure end to end — application architecture, APIs, frontend, AWS deployment, realtime systems, identity, storage, and offer workflows. The product has been sunset.

  • Designed and implemented the product and infrastructure: React client, Node.js APIs, PostgreSQL, Dockerized services on AWS EC2, S3 storage, email, SSO, KYC, maps, and Socket.IO realtime updates
  • Reached approximately $25M in aggregate value of active property listings at peak
  • Designed OfferFlow after conversations with realtors and people who had purchased multiple properties, then implemented offer creation, negotiation, and contract generation
  • Shipped a paid listing-promotion plan for real estate developers so their listings appeared above ordinary listings, and operated it for a small set of paying customers
  • Built an API release path that started the new version on a different port, validated it, then switched traffic — so API releases did not require downtime
Technologies
ReactRecoilNode.jsExpressPostgreSQLSocket.IODockerAWS EC2S3StripeMapbox

  1. Mar 2022 — Aug 2024

    Los Angeles, CA

    PsyFi

    Senior Full Stack Engineer

    Core product engineer for PsyFi and the PsyOptions protocol — built the primary surfaces for options trading, token launches, liquidity management, airdrops, staking, and mobile lending across the Solana DeFi stack.

    Highlights (6)
    • Built and shipped core PsyOptions trading experiences — American-style options on Solana including minting, collateral management, and order flow UX
    • Designed and built Fusion — collateralized options airdrops using Merkle trees for claim verification at protocol scale
    • Led frontend development of Armada — LBP curve auctions for token launches and CLMM vaults built on Orca Whirlpools; CLMM vaults reached over $1M TVL without marketing
    • Built a white-label staking frontend used by Solana teams including Bonk; partner deployments reached approximately $25M in staked assets
    • Led DeMUX — tokenomics revenue allocation tooling used by multiple Solana teams
    • Ported PsyLend borrow/lend flows to React Native days after Solana Mobile launch (Pyth / Switchboard oracles, DOV collateral)
    Technologies
    TypeScriptReactNext.jsReact NativeRustAnchorSolanaPsyOptionsOrca Whirlpools

  2. Aug 2020 — Mar 2022

    San Francisco, CA

    OnTraccr

    Full Stack Engineer

    First engineer on a SaaS platform for construction contractors — labor tracking, forecasting, data sharing, and automated workflows.

    Highlights (3)
    • Built timecard management with dynamic OT and pay calculation
    • Designed and implemented end-to-end multi-stage timecard approval plus budget and payroll analysis — core product features
    • Developed core React Native mobile features including dynamic forms, guides, and document center
    Technologies
    ReactReduxReact NativeNode.jsHapiMySQLAWSSocket.io

  3. Oct 2020 — Jan 2021

    San Francisco, CA

    Atlas One

    Full Stack Engineer (Contract)

    Research portal for asset-backed digital securities with live pricing and a composite index.

    Highlights (2)
    • Designed and developed the research portal for digital securities offerings
    • Built cron jobs that pull live price data from issuers and calculate an index from historical token price changes
    Technologies
    ReactReduxNode.jsExpressKeystonePostgreSQLGraphQLCircleCI

  4. May 2020 — Aug 2020

    San Francisco, CA

    WizeHire

    Software Engineer (Contract)

    Hiring platform work spanning customer support workflows and automated engineer onboarding infrastructure.

    Highlights (2)
    • Built job-posting management components that sped up customer and support workflows
    • Automated VM development stack — reduced new-engineer environment setup from about a week to a few minutes
    Technologies
    ReactDjangoPHPJenkinsGCPBash

  5. May 2019 — May 2020

    Vancouver, BC

    Philips

    Software Developer

    Distributed smart senior-home system — IoT device integration, event handling, and facility monitoring dashboards.

    Highlights (3)
    • Integrated third-party IoT devices and built apps to manage them and handle events
    • Built responsive dashboard components for monitoring device status across facilities
    • Mentored interns and new hires in a team of about ten while helping scale the product
    Technologies
    Node.jsAngularTypeScriptPythonKubernetesRabbitMQMySQLDynamoDBAWS S3

  6. Sep 2017 — Aug 2018

    Vancouver, BC

    Terramera

    Software Engineer

    AgTech engineering across data pipelines, internal analytics dashboards, plant imaging hardware software, and computer vision.

    Highlights (3)
    • Built cron-activated inventory download jobs across major retailers into PostgreSQL, saving hours of weekly manual work for sales analysts
    • Developed statistical and administrative tools for an internal dashboard used by sales and chemistry research teams
    • Trained a stained-egg segmentation model with 96% accuracy
    Technologies
    Vue.jsD3.jsJavaPostgreSQLDockerPythonKerasOpenCVMQTT

  7. Jul 2016 — Aug 2016

    Minsk, Belarus

    IBA Group

    Java Developer (Summer Internship)

    Android app for a supermarket network that scans passports via OCR to fill rewards documents.

    Highlights (1)
    • Built Android OCR flow using Google Vision and Tesseract for passport scanning
    Technologies
    JavaAndroidMaven

Selected projects

Flagship case studies

Five pieces of work first: transaction infrastructure, team product delivery, partner integration, analytics-led prioritization, and end-to-end product engineering. Additional work lives in the other filters.

6 deprecated or sunset · 6 still live

Skills

Technical capabilities

Grouped by how the work happens — individual technologies appear only when supported by experience.

  • Product engineering

    • TypeScript
    • React
    • Frontend architecture
    • Backend systems and APIs
    • Mobile applications
    • Dashboards
    • User experience implementation

  • Systems and infrastructure

    • Distributed systems
    • Cloud infrastructure (AWS, GCP)
    • Databases (PostgreSQL, MySQL, DynamoDB)
    • Messaging (RabbitMQ, Socket.io)
    • Containers and Kubernetes
    • Deployment and CI/CD
    • Reliability-minded delivery

  • Solana and decentralized finance

    • Solana transactions and modern client APIs
    • Public SDKs (@orca-so/tx-sender, orca_tx_sender)
    • Priority fees, Jito tips, compute estimation, confirmation
    • Integrations and partner tooling
    • Liquidity products and CLMM workflows
    • DeFi application development

  • Data and product analytics

    • Mixpanel
    • Event instrumentation
    • Funnels and cohorts
    • Retention analysis
    • Dashboard development
    • Using product data to change priorities

  • AI tooling in engineering

    • Internal AI-assisted code review
    • Test scaffolding and TDD with human validation
    • Model evaluation (reasoning, coding quality, reliability, latency, cost)
    • Prompt and context control, including how much diff or repository context is sent
    • Using smaller models when a larger one is unnecessary
    • Known failure modes: overfit tests, missed races, hallucinated APIs, shallow review

  • Leadership and product delivery

    • Engineering management (3 engineers + designer)
    • Planning, prioritization, and roadmap execution
    • Hiring and interviewing
    • Coaching and performance feedback
    • Technical and product risk management
    • Partner alignment and cross-functional delivery
    • Remaining involved in architecture and critical implementation

AI-assisted engineering workflows

Internal review, TDD, and model evaluation at Orca

I treat models as an engineering system: they need evaluation, constraints, feedback loops, and cost discipline. This is not machine learning research, model training, or production inference infrastructure.

Read the case study
Development problem
Product surface and partner deliverables grew faster than headcount. Test writing and pull request review were the bottlenecks, not idea generation.
Workflow before AI
Engineers wrote tests after implementation. Pull request review used ad hoc checklists and whatever the reviewer remembered about edge cases.
Information supplied to models
Module boundaries, existing test patterns, acceptance criteria, and the diff under review — not the entire repository by default.
How context was structured
Repo conventions, related test files, interface contracts, and task constraints bundled into structured prompts. Context size was an explicit choice: enough for the task, not a dump of the tree.
What models generated
Unit test scaffolds, edge-case suggestions, and pull request comments highlighting risk, missing coverage, and consistency issues. Output was never merged unreviewed.
Testing and review
Generated tests run in CI. Engineers review assertions, fixtures, and async behavior before merge. Review comments are checked against the actual diff. I used models to supplement human review, not to replace it.
Model selection
I evaluate new models as they become available and keep one only when it improves reasoning quality, coding effectiveness, reliability, latency, or cost. Difficult tasks escalate to a more capable model; routine scaffolding stays on a smaller or cheaper one. Model choice is intentional, not 'whatever shipped this week'.
Cost and context discipline
Unnecessary tokens are a defect. I reduce context, trim diffs, avoid resending files the model does not need, and prefer a cheaper model when the task does not require a stronger one.
Failure modes observed
Tests that overfit implementation details; missed asynchronous race conditions; hallucinated APIs; confident but wrong assumptions about system behavior; review comments that sound complete and miss architectural issues. Those failure modes are why human merge approval stayed mandatory.
What humans remained responsible for
Architecture decisions, merge approval, production behavior, and final judgment on test quality and review findings.
Observed result
The team used the workflow for TDD assistance and PR review. Coverage on new features and partner integrations arrived faster; the review bar stayed a human decision. The useful part was the evaluation loop — including knowing when the model was wrong — not the novelty of using an LLM.
Additional experience
  • Node.js / Express / Hapi
  • Next.js
  • React Native
  • Redux / Recoil
  • Vue.js
  • Angular / AngularJS
  • Django
  • PHP
  • Rust / Anchor
  • Python / Keras / OpenCV
  • Java / Android
  • GraphQL
  • D3.js
  • Docker
  • MQTT
  • Jenkins / CircleCI / GitHub Actions