Maxim Ignatev

Built AI-assisted contract authoring and document-analysis workflows for a multi-tenant legal AI platform.

• Built a contract-generation agent and TipTap editing interface with streamed documents, selection-based revisions, structured HTML operations and autosave.
• Engineered exact, trigram-based fuzzy and partial document search that resolves and highlights source passages even when LLM citations are paraphrased or slightly inaccurate.
• Designed and implemented a production-oriented prototype of a Python/PostgreSQL/Redis pre-queue scheduler with hierarchical fair-share allocation across tenants, users and workload classes, preventing large batches from monopolizing Inngest workers.
• Provisioned self-hosted Inngest with Terraform on Azure Container Apps, including KEDA autoscaling from 100 to 400 workers, private PostgreSQL/Redis infrastructure, replica-aware capacity tracking, CI/CD and observability.

• Added batched dispatch, Redis in-flight counters, round-robin job selection, idempotent completion and recovery of jobs interrupted by scheduler or worker failures.
• Integrated the scheduler with an existing grid document-processing workflow in a backend feature branch; the architecture remained a prototype and was not adopted into the main production flow.

Delivered short-scope contract work on existing React and Node.js codebases through Upwork.

• Implemented frontend features, third-party integrations and bug fixes across client projects.

• Scoped briefs directly with non-technical clients and adapted quickly to unfamiliar codebases.

Built the frontend from scratch for a collaborative user research platform: transcript editor, multi-tenancy, roles management and billing.

• Engineered a TipTap transcript editor with timestamp-preserving edits, speaker controls, media sync, autosave and offline support.
• Built the multi-tenant frontend covering workspace isolation, roles and billing.
• Owned frontend development as the first frontend engineer, taking the product from an empty repository through completion of the agreed frontend scope.

• Collaborated daily with design and backend in Agile product iterations.

Senior engineer on a genealogy platform, working on the media gallery, an AI video-story generator and platform-wide improvements.

• Built the media gallery with photo restoration, document OCR, audio transcription and translation workflows.
• Developed "MemoryClips", turning GEDCOM files into video stories with Claude, Azure Text to Speech and Remotion.
• Implemented serverless video rendering with Remotion on AWS Lambda and S3.

• Shipped UI/UX and product improvements across the existing genealogy platform.

Co-founded an AI product that extracted short-form clips from long videos and distributed them across social platforms.

• Built a media pipeline for transforming long videos into short-form clips.
• Developed a working prototype with video segmentation and an in-browser editing interface.
• Implemented a backend video-processing service using FFmpeg and AI-assisted title generation.

• The founding team discontinued the project at prototype stage after reviewing the competitive landscape.

Developed a specialized text editor for a legal case organisation and management product.

• Built TipTap-based workflows for capturing and organizing case facts.
• Implemented note management, drag-and-drop reordering and domain-specific handling for legal entities.

• Collaborated with the founder and a designer in weekly product iterations.

Built the end-to-end no-code automation for a Covid-19 testing telemedicine service.

• Designed and implemented the full order-to-result automation workflow.
• Integrated Shopify, Zoho CRM, Zoho Bookings and Mailgun through Zapier, covering order capture, scheduling, test-tube QR codes and PCR result delivery.

• Documented the workflow and prepared handover instructions for staff.

Co-founded and built a chapter-by-chapter publishing platform for writers, released briefly before the team pivoted the product.

• Built the platform from scratch as the sole technical founder.
• Implemented chapter-by-chapter publishing, likes, comments, notifications, donations and a followers system.

• Integrated Stripe donations and email notifications for new chapters.
• After a short release and a pivot, the founders ended the project over different product directions.

Developed a mobile app and PWA of resistance-band fitness tutorials.

• Developed and released iOS and Android apps using Bravo Studio.

• Integrated DatoCMS so the team could manage workout content without developer involvement.
• Built a Progressive Web App for browser-based access to the same content.

Client engagement through Web-pal focused on React dashboard and React Native mobile development.

• Developed various dashboard features using React and Redux.
• Initiated and drove development of the mobile application using React Native.
• Mentored junior developers on the team.

Client engagement through Web-pal building and shipping a React Native mobile application.

• Implemented a React Native interface from design specifications.
• Integrated backend APIs into a fully functional product.
• Published the mobile application to the App Store and Google Play.

Client engagement through Web-pal implementing responsive React and Redux frontend features.

• Developed frontend features using React and Redux.
• Implemented responsive designs using SCSS.
• Worked closely with backend developers to integrate new features.

Software agency role spanning outsourced client work and internal products, alongside mentoring and team culture.

• Progressed from junior to middle-level engineer, then took on leadership duties and mentored a junior developer.
• Delivered client engagements and 2 internal products (DBGlass, Chronos), all in Agile teams with peer code review.
• Shipped React Native applications to the App Store and Google Play.

Built a RAG chatbot that answers questions against your own curated library — YouTube videos, articles, tweets, books, social posts.

• Built an ingestion workflow for uploading, transcribing, segmenting and vectorizing mixed-media sources with Deepgram, OpenAI embeddings and Pinecone.
• Implemented a streaming chat interface with citations to exact positions in the source media.

Stopped: I was uncertain about the market and it would have consumed resources I did not have to spend on an experiment.

Built a platform to "speed-watch" YouTube videos by segmenting them into chapters and key takeaways, synchronizing transcripts with playback and ranking passages by relevance. Added a claim checker that looked up statements through the Perplexity API.

I thought the claim checker might help push back on populism.

Stopped: it does not solve a real problem — it mostly lets you procrastinate faster. If a product does not solve a problem, it is doubtful people will pay for it, and I could not risk more resources on it.

Built a Chrome extension that summarized highlighted text in place on a double-tap of Enter, using a DistilBERT model through the Hugging Face API.

Stopped: waiting five or more seconds for a summary (as it was at the time) was slower than just scanning the text diagonally yourself.

Built a parametric design-system prototype that retuned UI components through five visual parameters: colors, fonts, spacing, borders and shadows.

The idea came from Tailwind UI, which gives you the building blocks for common site types but leaves them looking like Tailwind UI.

Stopped: I pitched it to investors without success, so I shelved it and went back to contract work.

A Chrome extension using the GPT-J model to bring text generation into any input field on any website.