Experience, systems, and proof of work.

Six-plus years building backend, financial, and applied-AI systems, with hands-on ownership from architecture through production.

Open full-screen ↗Download PDF ↓

Document preview

Pages 01—02

If the embedded viewer is unavailable on your device, open the PDF full-screen or download a copy using the controls above.

Accessible text version

Semantic HTML résumé

Manuel Parra

Lead Software Engineer (IC), Python Financial Data Pipelines, AI/RAG & Multi-Tenant Platform Architecture

Phone: +58 424 434 8349Email: th3nolo@gmail.comNationality: VenezuelanLinkedIn: linkedin.com/in/th3noloWebsite: th3nolo.comGitHub: github.com/th3nolo

Professional summary

Lead-level software engineer and backend/AI data platform engineer with 6+ years of professional experience, plus earlier product-building and hackathon work. Specializes in Python/FastAPI, financial data ingestion pipelines, SQL data modeling, OAuth/webhook integrations, AI/LLM and RAG workflows, and Azure/GCP/AWS cloud delivery. Experienced translating client and product requirements into tenant-aware platform architecture, normalized data models, external financial integrations, and production-ready systems while remaining hands-on as an IC.

Technical skills

Backend & Architecture: Python, FastAPI, Node.js, TypeScript, NestJS, Django, Rust, API design, software architecture, service boundaries, tenant-aware access
Financial Data & Integrations: Financial data ingestion, P&L and income statements, balance sheets, cash flow, EBITDA, Plaid OAuth, Bridge.xyz, KYC/KYB, virtual accounts, webhooks
Data & Databases: PostgreSQL, SQL, Azure SQL, MongoDB, Redis, Pydantic, normalized schemas, financial-statement mapping, source evidence, benchmarking
AI & Data Systems: OCR, Document AI, RAG pipelines, multimodal and vision LLMs, PydanticAI, Langfuse, OpenAI API, Azure OpenAI, Hugging Face
Cloud, DevOps & Compliance: Azure, GCP, AWS, Docker, Podman, CI/CD, GitHub Actions, Railway, Linux, audit logging, access control, HIPAA-aware patient-record handling familiarity
Fintech, Web3 & Automation: Plaid, Bridge.xyz, Aave V3, Privy, Solidity, Foundry, virtual accounts, compliance-state sync, Hedera/Hiero

Professional experience

Technical Lead & Full-Stack Platform Architect (IC)

ConTrust Suite — Construction Daily Reporting Platform

March 2026 — Present

Architected a tenant-aware, project-scoped construction daily reporting platform for jobsite document ingestion, SharePoint-synced files, dashboard analytics, and AI-assisted reporting.
Implemented source-traceable ingestion and canonical mapping flows that convert construction daily-report PDFs and SharePoint files into normalized reporting data with raw values, validation flags, source references, and document evidence.
Orchestrated FastAPI backend services, React/Vite workflows, PDF extraction workers, shared Pydantic schemas, extraction benchmarking, and AI reporting modules across a multi-repo platform.
Strengthened extraction reliability across five pipeline controls: source-evidence contracts, idempotent worker completion, real-document regression tests, accuracy benchmarking, and retry/concurrency handling.
Delivered seven product-facing workflows across source-file review, report navigation, dashboard metric mapping, missing-data semantics, canonical review, manpower and equipment analytics, and grounded daily briefs and Q&A.

Technical Lead & Full-Stack Product Architect (IC)

ConTrust (Remote)

January 2026 — March 2026

Spearheaded a three-engineer build of ConTrust, a contractor prequalification, KYB verification, and financial-risk platform for the Saudi construction market.
Authored five architecture documents and 44+ product requirements from client and product conversations, converting requirements into architecture decisions, domain rules, and delivery priorities.
Improved financial Document AI extraction from 34% to 94% accuracy on real audited statements, processing a 22-page filing in 97 seconds at roughly $0.30–$0.42 per company.
Modeled PostgreSQL/FastAPI domain layers spanning company onboarding, KYB/UBO verification, canonical fact and ratio mapping, score reports, credit wallet and payment flows, supply-chain scoring, audit logs, and event-driven notifications.
Composed an agentic Document AI pipeline for Arabic and English KYB and financial PDFs, combining OCR, vision LLMs, layout-aware extraction, deterministic validation, source grounding, and human review.
Mapped financial-statement extraction across income statements, balance sheets, cash flow, EBITDA, liquidity, leverage, and KYB risk indicators for downstream financial-risk reporting.

Systems Architect, Rust Cryptography & Consent Infrastructure (IC)

LayerD (Remote)

October 2025 — December 2025

Engineered the Rust cryptographic core for identity-based encryption, BLS signatures, Merkle proofs, threshold decryption, PVSS/DKG flows, and WASM/FFI exports.
Verified 3.63M real device IDs in 63 seconds through hashed consent records, Merkle batching, Hedera anchoring, and multi-node proof verification.
Secured gateway verification flows for consent ingestion, Merkle batching, HCS anchoring, proof indexing, provider isolation, OIDC/JWKS access tokens, API-key rotation, and restart-safe proof recovery.
Deployed threshold-node infrastructure with NATS JetStream transport, DKG convergence handling, encrypted key-share storage, deterministic coordination, health checks, metrics, and Railway/Podman hardening.

Backend Engineer & Systems Architect (IC)

Stackit.ai (Remote)

March 2024 — October 2025

Owned backend and platform architecture for StackIT's onchain wealth automation product, connecting smart-wallet infrastructure, Aave V3 health-factor protection, Bridge.xyz/Plaid onboarding rails, Privy authentication, scheduled workers, admin operations, and frontend execution workflows.
Connected Plaid OAuth and Bridge.xyz rails for KYC/KYB onboarding, bank linking, virtual accounts, transfers, static memos, webhooks, idempotency, and compliance-state synchronization.
Processed $500K+ through Bridge/Plaid-linked virtual accounts over six months with no integration incidents.
Automated Aave V3 health-factor protection across 84 smart-wallet positions, supporting $2M+ collateral flow and zero liquidations over 18 months.
Defined StackIT AI v2 architecture for wallet automation, external-agent gateway APIs, simulation-gated execution, policy enforcement, webhook lifecycle delivery, safe mode, and Base deployment readiness.

Backend Engineer, Real-Time Sports Gaming Platforms (IC)

Zingiber (Remote)

September 2020 — March 2024

Built foundational backend systems for real-time fantasy sports and gaming products, owning API design, MongoDB schemas, WebSocket state flows, sports-data integrations, debugging, and documentation.
Created REST APIs with Express, MongoDB, Swagger/OpenAPI documentation, validation, error handling, CORS, environment configuration, and deployment setup.
Synchronized five-player lobby mechanics with Socket.IO, supporting public and private squads, membership updates, lobby joining, player selections, and shared game state.
Linked external basketball and sports APIs to fantasy mechanics around teams, scores, fixtures, and user predictions.
Contributed to Regalis, a hackathon blockchain prototype using Solidity, Hardhat, Chainlink price feeds, IPFS metadata, and Mumbai deployment scripts for donation-based NFTs.

Education

Bachelor's Degree in Computer Science
Universidad de Carabobo · 2013 — 2018

Languages

Spanish — Native
English — C1

Awards & recognition

Moralis 1st Place Winner, Chainlink Fall Hackathon 2021 — awarded USD 10,000 for Regalis.
Moralis Pooled Prize Winner, Chainlink Spring 2022 Hackathon — winner for Monkey Page Lab in the USD 13,500 pooled prize category.