Fractional Safety Management

Keep the signature. Hand us the system.

The FAA requires a Part 135 operator to keep certain roles internal — a signatory, an Accountable Executive with real authority. Everything else an SMS actually requires — building it, running it, reading it, keeping it alive between audits — does not have to live inside your operation at all.

We build and run your SMS end to end — SRM, safety assurance, safety promotion, records, reporting. Your reports come to us, and we read them alongside every other operator's — separately, never merged — because the risks you're flying through rarely stop at your fleet. You sign. We run.

After 28 May 2027, the distinction that matters isn't whose manual reads best. It's who can produce the record on request — for an inspector, an underwriter pricing your policy, or a lender's risk desk. That's the line this page is built around.

At a Glance
Deadline
28 May 2027 — full SMS compliance, FAA Part 5
Applies to
All Part 135 certificate holders — charter, commuter, air taxi
We run
Safety Policy (co-authored) · SRM · Safety Assurance · Safety Promotion — every Part 5 pillar
Stays with you
Signatory authority · Accountable Executive — non-delegable by regulation
›Decouple. ›Act. ›Observe. ›Fix.
01

Why Now

Every Part 135 charter and commuter operator — not just the large ones — must declare a compliant Safety Management System by 28 May 2027. A functioning SMS has to be demonstrated in operation, not assembled the month before, and most operators this size have never run one and don't have the headcount to stand up a safety department just to satisfy a filing.

A template manual satisfies the letter of that requirement and misses the point of it entirely. An SMS nobody reads is a dial with nothing behind it — the audit passed, the binder complete, and no one who would notice the beat lengthening. The FAA doesn't grade the binder. It grades whether the record it asks for matches the manual — and increasingly, so do the insurers and lenders underwriting the risk behind it.

Drift Watch
Public Health · 2020–2021

COVID

Every operator faced the same virus and answered it alone. Quarantine rules, cabin sanitization procedures, and crew rest adjustments were drafted independently, then revised again as guidance shifted, sometimes contradicting the operator down the ramp. Fatigue and procedural-deviation reports moved at rates that differed sharply operator to operator, with no shared baseline for what counted as normal. A risk framework built around one operator's own numbers has no way to tell an adaptation from a drift.

Labor · Oct–Nov 2025

The shutdown

A 43-day lapse in federal funding left roughly 13,000 air traffic controllers working without pay. Control towers ran understaffed or at reduced capacity nationwide. Confidential safety reports, including submissions to NASA's Aviation Safety Reporting System, rose over the same period. With no shared mitigation plan, individual operators designed their own workarounds, and decentralized, sometimes conflicting procedures spread across the national airspace system. No single operator's risk framework is built to register a pattern at that scale — which is why it went unseen for weeks.

Trade Policy · 2025–2026

Tariffs

Tariffs of 25% on steel and aluminum, introduced in 2025, added an estimated $5 billion a year in industry-wide production costs. Individual manufacturers absorbed hundreds of millions of dollars each and responded by qualifying new suppliers under time pressure, independently, with no shared data on how those substitute parts were performing. Whether the substitutions hold up over time isn't known yet — and no operator's own risk framework is built to find out before it matters, industry-wide.

Maintenance · 2025

Supply chain

IATA estimates airlines absorbed more than $11 billion in unplanned costs from supply chain disruption in 2025 alone, as the industry's aircraft order backlog passed 17,000. Facing parts and MRO delays, individual operators extended maintenance intervals and deferred non-critical work, each under its own internal pressure, none checked against a shared standard. No record exists of how many operators were deferring the same category of work at the same time — only that each one believed it was the exception.

Components · 2025–2026

Chip shortage

A persistent semiconductor shortage has delayed avionics deliveries since the pandemic, and operators have responded by deferring minimum equipment list items and substituting components as parts become available. Each deferral is evaluated and logged inside a single operator's own maintenance program. No registry tracks which substitutions have held up in service and which haven't, industry-wide. A workaround proven safe at one operator is invisible to the next one making the identical call, blind.

Weather · Winter 2025–26

La Niña

A weak La Niña shaped the winter of 2025–26: heavier snow and cold in the north, drought and dry, turbulence-prone air in the south. De-ice planning and routing decisions were made operator by operator, each against its own historical baseline, with no reference to how peers were handling the identical weather pattern. Every adjustment read, locally, as routine seasonal judgment. Whether any operator's risk tolerance had quietly shifted from the last comparable winter wasn't something any one operator's framework was built to notice.

Energy · Feb–May 2026

Jet fuel doubles

A Middle East conflict closed the Strait of Hormuz in February 2026, pulling 10 to 15 million barrels a day off the global market. Jet fuel prices more than doubled within ten weeks. Individual carriers revised fuel-reserve and alternate-planning margins under cost pressure, each against its own historical baseline, none checked against the others. Every revision looked, internally, like ordinary cost discipline. None of it was compared across operators, so no one could tell prudent trimming from a fleet-wide erosion of margin.

Analysis

The pattern is drift

None of the seven incidents above were unique to one operator. Each read locally as routine, because every operator's own risk framework, measured only against itself, had no way to register the pattern. That's not a data gap. It's the operating condition of a decentralized system — and it's exactly what a cross-operator practice is built to see first.

02

What Stays. What Moves.

The smallest internal footprint the regulation allows, and nothing smaller than that.

Stays with you

What the FAA won't let leave

The certificate holder's signatory authority, and an Accountable Executive with real, demonstrable control over the operation. These cannot be delegated to an outside firm under any arrangement, and we don't pretend otherwise.

Runs through us

Every pillar Part 5 requires

  • Safety Risk Management — hazard identification, risk assessment
  • Safety Assurance — audits, internal evaluation, corrective action tracking
  • Safety Promotion — training content, communication, culture work
  • Hazard reporting intake, records, and the Declaration of Compliance package
  • Reading the record — continuously, not once a quarter

Exactly which roles are legally irreducible depends on your certificate's size and structure. We confirm the precise split with you and your counsel before anything is signed — the principle above is the target, not a guess.

03

How It Works

Decentralized ownership. Centralized operation.

I

We build the system

Not a template. An SMS structured around your actual operation — your aircraft, your routes, your existing records — because that's what the regulation actually requires and what a boilerplate manual only pretends to satisfy.

II

We run it

Day to day, not just at audit time. Reports come to us. We administer the hazard register, the risk assessments, the training cadence, the recordkeeping — the operational labor of an SMS, carried by us instead of built out in-house.

Sample Finding — De-identified
SourceHazard report, ground operations
SignalThree fuel-quantity discrepancies logged as resolved in six weeks — each closed by the reporting crew member, none cross-referenced against the others.
FindingNot a fueling error. A reporting habit: the same person opening and closing the loop on their own report, with no second reader before the record closed.
ActionSecond-reader sign-off added before closure. Pattern flagged for the fleet's next 90-day trend review.

This is what reading the record instead of the dial produces in practice — a pattern caught before it needed a citation to be taken seriously.

III

Many operators. One environment.

Weather patterns, geopolitical risk, seasonal surges. Every pilot is flying through the same environment, but each operator only ever sees their own slice of it — logged in one hazard register, compared to nothing. Reading across every operator we run means those signals surface before they'd show up in any single operation's record.

IV

You sign what's true

Your Accountable Executive reviews and signs; we've already done the reading. The declaration of compliance reflects a system that has actually been run, not assembled the week before it was due.

04

The Engagements

Not a compliance shop. No template manual dressed up as an SMS. Two ways to start, sized to where your operation actually is.

›

The BuildEntry point

Gap analysis against Part 5, and an SMS built for your operation from the ground up. We read your existing records, your near-misses, your quiet write-arounds before a page of the manual gets written — so what we build reflects an operation that has actually been read, structured to hand off into ongoing operation rather than sit in a binder.

Timeline
Typically 6–10 weeks — scoped to fleet size and existing documentation
Deliverables
SMS manual · hazard register, built and populated · training curriculum · draft Declaration of Compliance
Fleet range
Single-aircraft operators through multi-aircraft fleets
Exact scope and timeline are confirmed after reviewing your certificate and current records — the range above is illustrative, not a quote.
›

Managed SMS OperationOngoing

We run your SMS as an outsourced operational function, not an advisory retainer. Reports route to us; we administer risk management, safety assurance, and safety promotion; we carry the cross-operator reading no single in-house hire can replicate — for a fraction of the cost of standing up a safety department to hit one deadline.

Cadence
Continuous intake · monthly reporting cycle · quarterly audit-readiness review
Point of contact
One person, across every FAA interaction — not a rotating team
Pricing
Structured to fleet size and complexity — confirmed before signing, not published here
05

Audit-Ready by Design

You don't prepare for the audit. You're already in it.

Most operators experience an audit as an event — weeks of reconstructing records, explaining gaps, and hoping the story holds together under an inspector's questions. That scramble is itself evidence of an SMS that exists more on paper than in practice.

Because we run your SMS continuously rather than assembling it before a deadline, alignment isn't something we build in the weeks before an evaluation. It's the condition your operation is already in.

I

One point of contact

We interface with the FAA directly. Initial and continuing evaluations, correspondence, corrective action — coordinated through one expert who already knows your operation's full record, so no one is translating between an inspector and internal staff who weren't in the room.

II

No pre-audit scramble

Audit prep isn't a special project. The record an inspector wants to see is the record we've already been keeping, in the ordinary course of running your system — not reconstructed the week the letter arrives.

III

We carry the pattern across every audit we sit through

Every evaluation teaches us what the current cycle of scrutiny is looking for. That knowledge moves into your next one.

We bring the headache, so the regulator doesn't have to come looking for it.

06

A Framework for Sharing Safety Information

Safety data should move freely between the people who need it. And nowhere else.

Your data is never pooled with another operator's. Cross-operator findings we share back with you are de-identified and pattern-level only — never another operator's records, never anything that could expose them competitively or legally. Any boundary beyond that is set by you before an engagement begins, in writing, not discovered afterward.

Part 5 already requires a confidential internal reporting channel inside every operator, because safety data only tells the truth when the person filing it isn't afraid of what happens next. That protection has no equivalent for information moving between operators — the layer a cross-operator practice like ours sits in. We think aviation safety data eventually needs its own version of what HIPAA did for medical records, and we're involved in early work on what that could look like — nothing here is settled law yet. Until it exists, contractual confidentiality is what does that job.

07

Who It's For

Part 135 charter and commuter operators who don't have — and don't want to build — a full-time safety department, but do have a federal deadline. Multi-aircraft operations where the chief pilot is also the safety manager is also the person flying the trip. Anyone who wants the smallest possible internal footprint and the most rigorously run system on the other side of it.

08

Behind It

Founder
Shaurye Chakravarty
Ratings
Commercial Pilot — ASEL, AMEL, Instrument · CFI / CFII / MEI
Research
M.Sc. candidate, Safety Science — Embry-Riddle Aeronautical University, Prescott
Published
CONSCIENCE: A Radical View on Safety and Accountability — the argument this practice applies at organizational scale

Read the book and the author's other work →

09

Start

Not a sales call — a first read of your record

Send the last three occurrence or hazard reports from your existing system — redact whatever you need to. We'll read them the way we'd read them as your Director of Safety and tell you plainly where the gaps are. No cost, no obligation, and no template pitch afterward.

May 2027 is closer than the paperwork makes it feel. If you'd rather just talk first, that's fine too.

Send your last three reports↗ Or just start a conversation↗

An operation that reads stable is not the same as an operation that is keeping time. The difference is a record you can produce — not one you have to explain.