Open
WorkDemosServices Case studiesProcessJourneyContact
Autonomous business systems · Karachi

Hashir Jamil*

Hashir Jamil builds software that does the work — it finds the lead on a public register, calls, qualifies and books the meeting on its own. Nine systems. Four live right now.

See the systems
The systems

One engineer. Nine systems that act — open any of them.

About

Most portfolios show interfaces. This one shows systems that act.

Real phone calls. Real email. Real orders with real money. That one difference drives every decision in the work — the safety gates, the circuit breakers, the fact that every system here can be stopped.

I also carry a quota. Five commercial roles across SaaS, fleet technology and agencies — currently Sales Lead at Design Pro Verse, running a team of four. The revenue platform that team sells into, I built.

And through the studio: the sites, apps, identity and the registration and trademark paperwork a business needs to exist. Two lines. Both answerable to a number.

0
Systems built
0
Live in production
0
Domains
0
Engineer

“Every past defect is now a rule the build enforces.”

Hashir Jamil
Hashir JamilKarachi, PakistanBuilding
since 2022
Selected work

Nine systems. Four of them are awake right now.

Open any one for the full record — the problem, the build, and the proof.

SYS-08 / 09

Synapse AI & USD Halo

Craft studies · 20 and 15 files

Two small front-end pieces, built on purpose. Synapse AI is a motion-led marketing surface with smooth-scroll choreography; USD Halo is a compact React prototype. They are where motion and layout technique get tested before it reaches anything that handles money.

Live demos

The consoles, running.

Six real interfaces, rebuilt faithfully and animated. Click through the rail. Nothing here is connected to anything — that is the point.

hashir@workstation / rifl-crm
H

Revenue operations · overnight briefing

Four channels — voice, email, SMS and scraping — deciding the next action and executing it. 53,498 lines, 414 files, 150+ endpoints.

Calls placed
0
▲ 12.4% vs yest.
Conversations held
0
▲ 8.1%
Meetings booked
0
▲ 21.0%
Opt-outs honoured
0
suppressed x4 ch.
Lead heat decay · 72h half-lifeauto-refresh
Agent call · in progressLive
00:47
AgentHi — is this a decent moment for two minutes?
ProspectI'm not available right now, honestly.
AgentUnderstood. Would Thursday morning work better?
ProspectThursday's fine. Send something over.
Classify→ warm callback · booked · not voicemail
Work queueprioritised by heat
Callback due — bad timingHEAT 94
Sequence step 3 queuedHEAT 71
Voicemail — retry windowHEAT 46
Opt-out — suppressedHELD
Compliance gatesenforced in code
Do-not-call registry
Regional calling windows
Cross-channel opt-out sync
Autonomy — propose only

Lead sourcing · USPTO public register

Searches the public trademark register, enriches each hit from TSDR, and streams contactable leads into the CRM. The front door to everything else.

Public record USPTO filings are published by the United States government. Owner, mark, filing date, status and contact details are all part of the public notice system — this is public-record research, not private data.
Leads this run
0
bounded per extraction
Filter dimensions
0
intent, not volume
Enrichment workers
0
bounded pool
Record deadline
0
hung filings abandoned
Live scan · results streamingScanning
SerialMarkStatusContact
Filtersbuying signal
No attorney only
No website detected
Recently abandoned
Filing statusRecency windowFiling-date range US stateInt'l classEntity type No attorneyWebsite detectionRegion
Session strategybrowser cost paid once
01Playwright warm-up
1 browser launch · challenge cleared
02Authenticated HTTP
session reused · thousands of fast searches
03TSDR enrichment
per-record XML · email, alt email, phone
Handoff to RIFL CRMcall variables
serial_number98/412,660
mark_statusLIVE
filing_date2026-07-22
agent opens by naming the filingREADY

Risk control plane · 8 venues

Layered circuit breakers sit in front of execution. 298 files, 25,704 real trades analysed, three independent breaker layers.

Dead-man's switch engaged — synthetic feed detected on 2 venues. All orders halted, resuming after 60s of verified real data.
Venues connected
0
priority routed
Trades analysed
0
best-hours model
Breaker layers
0
all armed
Exposure used
0
of hard limit
Executed volume by hourUTC
Order tapepre-execution checked
Venue connectivityregistry
VEN-01
12ms
VEN-02
18ms
VEN-03
9ms
VEN-04
31ms
VEN-05
14ms
VEN-06
22ms
VEN-07
27ms
VEN-08
16ms
Circuit breakersindependent layers
Startup validation
Data-integrity monitor
Per-exchange limits
Dead-man's switch

Autonomous sourcing · scheduler

Ingests supplier stocklists in inconsistent formats, matches and de-duplicates, then runs outbound on a schedule. Boots fully functional with zero credentials.

Files
0
9,673 LOC
Scheduled jobs
0
unattended from boot
Integrations
0
full offline mock
Rows matched
0
de-duplicated
Job runner8 jobs
Stocklist ingest · 4 suppliers0%
Inventory de-duplication0%
Demand matching0%
Outbound campaign · 06:00 UTC0%
Telemetry export0%
Reasoning tiergraceful degradation
01Frontier model
billing failure · cooldown 14:22
02Local model
CPU inference · serving now
03Deterministic heuristics
standby · always available
Stocklist ingestion
supplier-04_stock_FINAL(2).xlsx

Workspace validation · monorepo

352 files across an apps-and-packages monorepo on a hosted cloud backend. Build, typecheck and test run workspace-wide from a single command.

Files
0
apps + packages
Packages
0
workspace linked
Commands
0
validates everything
Schema drift
0
clean
pnpm -w validate0.0s
apps/web · build
packages/ui · typecheck
packages/db · migrate
packages/auth · test
packages/config · lint
schema · diff
Package build timescold cache
Managed auth + PostgresHOSTED
Scripted reset & diff toolingREADY

Render queue · 7 repositories

One operator interface governing a render engine, a CLI, agent skills, a plugin, launch tooling, a deployment template and a live streamer. 23 files that coordinate rather than absorb.

Repositories
0
one interface
Agent skills
0
loaded
Control files
0
deliberately thin
Pipeline
HTML→MP4
with synthesised speech
Render jobs dispatchedvariant sweep
v01
v02
v03
v04
v05
v06
v07
v08
00:01RENDERhtml → framesv01
00:04TTSspeech synthesisedv01
00:09MUXmp4 written · 1080pv01
00:11RENDERhtml → framesv02
00:15TTSspeech synthesisedv02
00:20MUXmp4 written · 1080pv02
00:23SKILL19 agent skills loadedok
00:01RENDERhtml → framesv01
00:04TTSspeech synthesisedv01
00:09MUXmp4 written · 1080pv01
00:11RENDERhtml → framesv02
00:15TTSspeech synthesisedv02
00:20MUXmp4 written · 1080pv02
00:23SKILL19 agent skills loadedok
Governed repositories7
render-engineREADY
agent-skills19 LOADED
go-cliBUILT
live-streamerIDLE
deploy-templateSYNCED
launch-toolingv2.1

Business setup · delivery pipeline

The studio line: identity, product surface, and the registration and trademark paperwork behind it — tracked through one workflow with the agency team.

Engagement · “Northgate Logistics”Stage 3 of 5
01
Discovery
02
Identity
03
Build
04
Registration
05
Trademark
Product surface · responsive builddesign → dev
Deliverables tracker11 services
Brand identity & logo0%
UI/UX design system0%
Website development0%
Mobile application0%
Business registration filing0%
Trademark assistance0%
Demonstration · faithful to the real interface · deliberately wired to nothing
Services

Two lines. Both answerable to a number.

One is me at a keyboard — agents, lead generation and the platforms around them. The other is the studio team, covering everything a business needs to exist and be seen.

A

Systems engineering

Autonomous software, built against production from day one.

Voice and workflow agents that hold a real conversation, interpret the outcome, take the next action, and escalate to a human only when they should. Every agent ships disabled and in propose-only mode; autonomy is enabled per feature, reversibly, once its proposals have been reviewed and found sound.

Voice agentsOutcome classificationEscalation rulesPropose-first

Pipelines that mine public registers for buying intent and turn them into contactable, enriched leads — no purchased lists, no data anyone else already has. Grab&Snap does this against the USPTO trademark register: a business that has just filed a mark is publicly announcing a brand launch, which is the strongest timing signal there is. The same method generalises to company formations, trade licences, building permits and planning applications.

Public-register miningContact enrichmentIntent filtersStraight into the CRM

End-to-end outbound engines that source and validate prospects, run multi-step sequences, track engagement, and follow up until a lead converts or opts out. Compliance is enforced in code — do-not-call honoured, regional calling windows respected, opt-outs suppressed across every channel simultaneously.

Multi-channelHeat scoringA/B sequencingCompliance gates

Internal platforms and dashboards that replace spreadsheets and manual process — built around the way the business actually runs, then operated in production. Not handed over and forgotten: instrumented, supervised, and on-call.

Next.jsNodePrismaOpenTelemetry

Go-to-market planning across inbound and outbound, owned to the number — pipeline design, qualification criteria, forecasting, and the reporting that keeps it honest. Written by someone who has carried the quota, not only the roadmap.

Pipeline designForecastingTerritory analysis

Building sales teams from the ground up: structured onboarding, skills training, performance management, and the enablement material that makes it repeatable. Currently leading a team of four at Design Pro Verse.

OnboardingEnablementPerformance mgmt
How an engagement runsdirect
1 · Operational auditweek 1
2 · Propose-only buildweeks 2–5
3 · Supervised autonomyweek 6+
4 · Operated in productionongoing

Nothing is handed over and forgotten. Every autonomous component ships disabled, is promoted per feature once its proposals have been reviewed, and stays instrumented and reversible after launch.

B

Studio & business services

Delivered with a team, for businesses that need the whole setup.

Website designEditorial, conversion-led interfaces
Website developmentFast, accessible, production builds
UI/UX designResearch, flows, design systems
Mobile applicationsiOS and Android product surfaces
BrandingPositioning, voice, visual identity
Logo designMarks built to survive scale
§
Business registrationIncorporation and filings, handled
®
Trademark assistanceSearch, filing and follow-through
Digital solutionsAutomation for the day-to-day
Business conciergeThe admin nobody wants to own
Technology consultingArchitecture and build decisions
Something elseDescribe it and I'll scope it
The studio line covers the work a business needs in order to exist and be seen — identity, product surface, and the registration and trademark paperwork behind it. Engagements run through the agency team; the systems line is mine directly.
Case studies

Four problems worth bragging about.

Feature lists are free. These are the problems that rewrote the architecture — and the rules that now stop them coming back.

Goal

Classify every call outcome accurately enough to act on it without a human reviewing the tape.

Challenge

Real prospects saying “I'm not available right now” were being marked as voicemail and dropped.

Technology

Realtime STT→LLM→TTS voice pipeline · WebRTC · Twilio · Next.js 16 · Prisma

Outcome

Bad timing is now a warm callback by invariant. Encoded as a permanent test case.

What was actually happening

Two causes compounded. The voicemail vocabulary shared phrases with live humans — greeting patterns, apologies, the word “right now”. And the learning loop was feeding those human phrases back into its own dictionary.

So the system got measurably worse over time, which is why the problem kept recurring after each apparent fix. Every patch treated the symptom; the loop rebuilt the fault within days.

The repair

Separate the two vocabularies so they can never contaminate each other, then encode a hard invariant above the classifier: bad timing is a warm callback, never a dead lead.

classify/outcome.tsinvariant
// before — one shared vocabulary, self-poisoning
vocab.voicemail.push(...transcript.phrases)

// after — isolated corpora + hard rule
const human = corpus.live       // never written to
const vm    = corpus.voicemail  // never written to

if (signals.badTiming)
  return { outcome: 'WARM_CALLBACK' }  // invariant
Real prospects stopped being dropped from the pipeline
Classifier accuracy stopped degrading over time
The invariant became a permanent regression test
The failure shape entered the dependency-free rule engine
Goal

Place real orders across eight venues without a human watching the feed.

Challenge

Feeds degrade silently to synthetic values; venues disconnect mid-session; minimums change without notice.

Technology

Node · Express · React · Vite · eight exchange connectors · 25,704-trade research set

Outcome

Three independent breaker layers. Confirmed live fills across multiple venues and assets.

What the documentation did not say

Every venue constraint was verified against live public APIs rather than trusted from documentation. That process surfaced a renamed trading pair whose minimum order size had silently grown past typical order size — an error that would have surfaced only as unexplained rejections in production.

The harder failure was subtler: market data does not stop, it degrades. Feeds fall back to interpolated or stale values that look entirely plausible to a naive strategy.

The repair

A dead-man's switch that halts all orders the moment live data degrades to synthetic values, and refuses to resume until 60 seconds of verified real data have passed. Plus startup validation and per-exchange limits as independent layers, so no single check is load-bearing.

risk/deadman.tslayer 3 of 3
if (feed.isSynthetic() || feed.staleMs > 2000) {
  halt('SYNTHETIC_FEED')
  resumeAfter = verifiedRealFor(60_000)
}
// layer 1 — startup validation
// layer 2 — data-integrity monitor
// layer 3 — per-exchange exposure limits
// no single layer is load-bearing
8 venues connected with priority-based routing
25,704 real trades analysed for strategy and best-hours selection
Automatic halt on data-quality degradation
Full audit trail with export for accounting
Goal

Build a fast, offline, searchable gallery over a design-prompt corpus.

Challenge

The supplied corpus was not what it claimed to be, and the preview imagery was missing entirely.

Technology

Node HTTP · vanilla SPA · zero dependencies · no build step · magic-byte validation

Outcome

9 files, 87 designs indexed, 85 previews recovered, nothing installed.

What the audit found

Four supplied repositories proved to be forks of one library. Two were byte-identical. The 132 files described only 87 distinct designs. And 44 of 45 “premium” designs were identical to their free counterparts.

Preview images did not ship at all. They were recovered by reading the source site's own JavaScript bundle — which embeds its catalogue — and matching entries by title, recovering 85 of 87.

The rule it produced

A file claiming to be video turned out to be an expired playlist. Media is now validated by magic bytes, never by extension or content-type header.

corpus/audit.logpre-build
$ audit ./repos
repos supplied .......... 4
byte-identical forks .... 2
files claimed ........... 132
distinct designs ........ 87
"premium" != free ....... 1 of 45
previews recovered ...... 85 / 87
dependencies installed .. 0
9 files, zero dependencies, zero build step
87 designs indexed; 85 previews recovered
Media validated by magic bytes, not by claim
Scope corrected before a line was written
Goal

Turn newly filed trademarks into contactable leads at the moment of strongest buying intent.

Challenge

The signal is public but unusable: two separate systems, a bot challenge, and contact details buried in per-record XML.

Technology

Python · Flask · Playwright · USPTO search index · TSDR document retrieval

Outcome

One browser launch, then thousands of fast searches. No purchased lists anywhere in the platform.

Why bought lists lose

A purchased list is stale, generic, and sold to your competitors the same afternoon. A business that has just filed a trademark is publicly announcing that it is launching a brand — the strongest timing signal available to anyone selling branding, web or launch services.

The obstacle was never legal, it was mechanical. The search index sits behind a challenge that plain HTTP requests fail. The obvious fix — drive a headless browser for every query — is ruinously slow at scan volumes.

The repair

Pay the browser cost exactly once. A single Playwright session is warmed at startup, clears the challenge, and hands its authenticated session to ordinary HTTP requests. One launch, then thousands of fast searches. Enrichment runs on a bounded six-worker pool with a hard seven-second deadline, so one hung filing can never stall a scan.

sourcing/session.pywarm once
# naive — a browser per query, unusable at scale
for kw in keywords: browser.goto(SEARCH, kw)

# actual — warm one session, reuse it
session = playwright_warm()   # 1 launch
for kw in keywords:
    hits = http.get(SEARCH, cookies=session)   # fast

# enrichment: bounded, deadlined, streamed
pool = Pool(6, deadline=7.0)   # hung record dropped
Two USPTO systems combined — search index plus TSDR
Contact details read from each filing's own XML, never guessed or bought
Nine filter dimensions targeting intent, including DIY filers with no attorney
Serial, status and filing date flow straight into the voice agent's call variables
Public record, not private data

USPTO filings are published by the United States government. Trademark registration is a public notice system — applicants file knowing the record is public. The same method generalises to any public register: company formations, trade licences, building permits, planning applications.

Process

How one person ships nine systems.

No team, no cloud build fleet, no GPU. The constraints were not incidental — they are what forced the architecture to be simple enough to reason about.

01
Observe

A real operational failure surfaces.

02
Measure

Instrument and reproduce before touching code.

03
Root-cause

Follow the failure past its symptom.

04
Fix + encode

Repair, then add a rule so it cannot return unnoticed.

05
Verify live

Confirm against the running system, not the suite alone.

01

Verification over trust

AI-assisted development, nothing trusted because it was generated. Every claim verified by running it.

02

Measurement over assumption

Diagnose by measurement, never by assumption. A “corrupted build cache” turned out to be memory starvation.

03

Mock-capable by default

Built against real providers, then given full offline mock modes. Nothing untestable because a vendor key is missing.

04

Adversarial review

A structured multi-team audit found three critical and six high-severity issues. All fixed and verified.

05

Defects become rules

A dependency-free rule engine scans for the shape of every past production bug and fails the build on a regression.

06

Safety before autonomy

Every autonomous component defaults to propose and ships disabled. Autonomy is enabled per feature, reversibly.

Journey

Five years selling it. Then I built it.

The sales career is not a footnote. It is why the engineering knows what is actually worth automating.

Present · Current

Sales Lead

Design Pro Verse · Creative & digital agency

Lead and mentor a team of four across onboarding, training and performance. Own the full funnel from prospecting to close, and design the go-to-market strategy behind it. Built the revenue platform the team runs on, in-house.

Team of 4Full-funnel ownershipBuilt the platform in-house
Previous

Account Executive

Motive Technologies · Fleet & ELD solutions

Owned a portfolio of enterprise clients running ELD hardware and fleet tooling — renewals, contract negotiation, technical diagnosis through the CRM, and proactive hardware replacement to protect uptime and compliance.

US SaaS enterpriseRetention ownershipTechnical diagnosis
Feb 2024 — Mar 2025

Team Lead

TGT Nexus · Web development & digital marketing

Set brand direction and market positioning across web and social verticals, coordinated a cross-functional delivery team, ran paid acquisition end to end, and built the internal sales enablement programme.

Owned brand & GTMPaid acquisitionEnablement
Apr 2023 — Jan 2024

Business Unit Head

Qtechnologies

Communicated risk to sponsors at every level, kept programme reporting accurate across all work streams, and managed deliverables and timelines alongside the architecture and delivery teams.

Sponsor reportingRisk communicationDelivery management
May 2022 — Mar 2023

Team Lead

Techigator

Consultative selling backed by cost-benefit and needs analysis, sustained high-volume outbound, and territory analysis with status reporting across the team.

20% above target50+ calls / day30+ leads identified
Capability & stack

What I actually reach for.

Autonomous agent designProduction · 3 systems
Sales automation & outbound engineeringProduction · flagship
Lead generation from public recordsProduction · USPTO register
Platform & API engineering150+ endpoints shipped
Risk & safety systems3 breaker layers · audited
Revenue strategy & GTM5 commercial roles
Team leadership & enablementTeam of 4
Front-end & motion craft2 dedicated studies
Depth is only claimed where something is running. Every number here was counted from source, not estimated — and anything not in continuous operation is labelled Built, never Live.
Technology stackin production
Application
Next.js 16ReactPython · FlaskViteTypeScriptNode 22ExpresstsxGo CLI
Data & infrastructure
PlaywrightUSPTO · TSDRPrismaSQLiteSupabasePostgresOpenTelemetryProcess supervisionMonorepo
Agents, voice & media
Realtime STT / TTSWebRTC voiceBarge-in & VADTwilioLocal LLM inferenceHTML→MP4 renderingFramer Motion
Commercial
CRM ownershipOutbound sequencingForecastingTerritory analysisContract negotiationEnablement
Achievements & measured results
20%
Above target
50+
Calls per day
30+
Leads identified
9
Systems built
63k+
Lines in production
9
Audit issues closed
63k+
Lines in production
Across four systems that run without anyone watching them.
20%
Above target
Carried a quota before building the tooling that carries it now.
100%
Measured, not estimated
Every figure on this page was counted from source in August 2026.
Contact

Tell me what the business can't do yet.

Hiring? I build the agent and own the number it produces — that combination is rare, and it is exactly what I do. Building? Systems, or the whole studio line: site, identity, product, registration, trademark.