Technical Project Manager

Technical Project Manager.Run delivery for 15–25 backend, frontend, and AI engineers building constructiontech and CAD drawing-automation software — with structural engineers on oneside of you and the codebase on the other.FULL-TIMEMANAGER · 7+ YRSEngineering — Delivery & Program ManagementTEAM YOU’LL RUN15–25 engineers · Backend · Frontend · AILOCATIONIndiaEXPERIENCE6–7 yrs min · 4+ yrs managing engineersWORK MODE HybridDOMAINENGAGEMENTAEC · CAD / BIM automation · Multi-tenant SaaSAbout BuildTwinBuildTwin is a multi-tenant SaaS platform that runs the day-to-day operations of constructionand infrastructure companies — their workforce, projects, commercials, and documents — on asingle system. Alongside it we build an AI drawing-automation platform that turns structuredelement data into production-ready technical drawings and exports them to PDF and DXF.Both products are built by a small, senior engineering group in India: backend engineers on anAWS serverless platform, frontend engineers on Angular and React, and AI engineers workingon drawing generation and data extraction. Sitting next to them — not in another building, notin another timezone — are the structural and detailing engineers who define what a correctdrawing actually is.That adjacency is our biggest advantage and our hardest coordination problem. This role existsto make it work.  Role overviewWe are hiring one Technical Project Manager to own delivery across our engineering teams inBengaluru. You will run 15–25 engineers across backend, frontend, and AI workstreams, andyou will be the person who turns requirements originating with our structural and detailingengineers into scoped, sequenced, shipped software.This is a technical role. You will not be handed a groomed backlog and asked to move cardsacross it. You will sit in the design discussion and have a view on how and when — you willpush back on an approach that will not scale, ask why a trade-off was made, and know thedifference between a two-day change and a two-sprint one. You are not expected to writeproduction code. You are expected to read it, understand the system, and never be the leastinformed person in a technical conversation you are chairing.The centre of gravity of this job is the floor, not the calendar. Most of your day isspent with the people building and the people specifying — closing the gap between what adomain expert means and what an engineer can implement deterministically. Statusreporting is a by-product of doing that well, not the job itself. What you’ll own. Delivery ownershipOwn the end-to-end delivery plan for every active workstream — scope, sequence,dependencies, capacity, and dates that actually hold.Break large, ambiguous product intent into engineering-sized increments with explicitacceptance criteria, agreed with the people who will build them.Run the operating cadence: planning, standups, refinement, demos, and retrospectives thatchange something the following week.Maintain one honest view of status — on track, at risk, slipped, and why — visible to thefounders and to the team at the same time.Manage cross-team dependencies between backend, frontend, AI, and drawing-automationworkstreams so no team idles waiting on another.Surface risk early, quantify it, and arrive with options — never an escalation without aproposal.Protect the team from thrash: absorb changing priorities, re-plan properly, and say no or notnow when a plan cannot take more.Drive releases to genuinely done — QA, UAT with domain experts, documentation, and postrelease verification included.Own and improve the delivery metrics that matter: cycle time, predictability of committedscope, defect escape rate, and rework.Run the release calendar across two products so neither becomes the permanent secondpriority. Technical partnershipParticipate substantively in technical discussions on design, sequencing, and trade-offs —and hold your own with senior engineers.Understand the architecture well enough to reason about blast radius: what a schemachange touches, why a migration is risky, why an API contract must be agreed before twoteams start.Challenge estimates that look wrong in either direction, and recognise when an estimate isuncertain because the problem is still unclear.Translate technical constraints into decisions leadership can act on, and business constraintsinto direction engineers can act on.Keep testing, observability, and debt-reduction work on the plan instead of losing to featuresevery single quarter.Insist that decisions get written down — architecture choices, drawing rules, API contracts,and the reasoning behind each.Stay close to the artefacts: read the tickets, the design docs, the PR titles, and the incidentwrite-ups.Chair technical discussions where the outcome is a decision with an owner and a date, not alonger discussion.